update the repo
This commit is contained in:
DatTT127
@@ -0,0 +1,215 @@
|
||||
# User-Research Result
|
||||
|
||||
⇒ DEFINING THE PROJECT-SCOPE:
|
||||
|
||||
Develop an interactive **`musculoskeletal care and education platform`** that transforms standard DICOM X-rays into a **`collaborative workspace`**. The system **`empowers clinicians with continuous education, diagnostic support, and objective clinical checks`**, while **`actively educating patients about their locomotive health to foster a true partnership throughout their treatment journey.`**
|
||||
|
||||
**⇒ End-user I have identified in this project (The Pilot-Project) scope shall include as follow:**
|
||||
|
||||
- The User Profile & User Persona ? - Given the User-Profile as below within the domain of healthcare in general (these are matter in NFRs)
|
||||
|
||||
|
||||
### User-Profile 1: Healthcare Senior Expert
|
||||
|
||||
- **Who they are:** PhD, Specialist II, senior department heads, professors, clinical directors, and national-level experts.
|
||||
- **Age range:** about 40–60+.
|
||||
- **Domain proficiency / responsibility:** highest domain depth; they own complex diagnosis, surgery, protocol setting, teaching, research, and escalation decisions.
|
||||
- **Working environment:** central or urban referral hospitals with the highest case complexity, the heaviest consult load, and the strongest pressure to standardize quality across many juniors.
|
||||
- **What they know:** specialty medicine, clinical standards, institutional policy, and how to manage hard cases with incomplete information.
|
||||
- **What they do not know:** ML model behavior, deployment constraints, validation methodology, and how to operationalize AI into workflow without disruption.
|
||||
- **Attitude toward AI/ML:** cautious but open if it improves safety, speed, or teaching quality; skeptical of black-box outputs and unproven demos.
|
||||
- **Pain points AI should help with:** triage prioritization, guideline consistency, overload reduction, second-opinion support, and research/teaching summarization.
|
||||
- **Hindrances in solution design:** liability concerns, low tolerance for workflow friction, demand for explainability, and strong resistance to tools that look like “extra work.”
|
||||
- **Role in Dev/R&D:** decision-maker, clinical validator, and pilot sponsor; not the best first user for UX discovery, but essential for approval and governance.
|
||||
- **How to approach them:** start from patient safety, evidence, efficiency, and institutional benefit; bring concise demos, local data, and clear failure-mode handling.
|
||||
|
||||
### User-Profile 2: Professional Clinician
|
||||
|
||||
- **Who they are:** MD, Master’s, Specialist I physicians, consultants, and mid-career clinicians.
|
||||
- **Age range:** about 30–45.
|
||||
- **Domain proficiency / responsibility:** strong clinical decision-making, specialty care, supervision of routine clinicians, and participation in departmental operations.
|
||||
- **Working environment:** major cities and provincial hospitals; high volume, mixed complexity, heavy outpatient/inpatient balance, and constant multitasking.
|
||||
- **What they know:** practical evidence-based medicine, standard workflows, referrals, diagnostic pathways, and how hospital systems actually behave.
|
||||
- **What they do not know:** deep ML mechanics, model drift, data governance nuance, and how to audit AI outputs in a rigorous way.
|
||||
- **Attitude toward AI/ML:** pragmatic and curious; they will adopt if it saves time and reduces repetitive work, but they quickly lose trust if the tool is unstable or slow.
|
||||
- **Pain points AI should help with:** chart review, note drafting, imaging/radiology support, guideline lookup, coding/admin burden, and triage support.
|
||||
- **Hindrances in solution design:** fragmented systems, poor interoperability, time pressure, incomplete data, and the need to justify decisions upward.
|
||||
- **Role in Dev/R&D:** power users, domain testers, and feedback bridge between leadership and frontline users.
|
||||
- **How to approach them:** focus on time saved, accuracy, integration with current workflow, and measurable outcomes; avoid abstract “AI transformation” language.
|
||||
|
||||
### User-Profile 3: Practitioner
|
||||
|
||||
- **Who they are:** college/intermediate-degree clinicians, nurses, assistants, and routine care providers.
|
||||
- **Age range:** about 22–35.
|
||||
- **Domain proficiency / responsibility:** routine care, procedures, monitoring, handoff, and protocol execution rather than independent complex diagnosis.
|
||||
- **Working environment:** provincial, district, and some commune-level settings; constrained staffing, high patient throughput, and fewer specialist backups.
|
||||
- **What they know:** operational care, standard procedures, basic triage, record keeping, medication handling, and patient communication.
|
||||
- **What they do not know:** advanced diagnostics, edge-case management, system design, or how to judge whether AI recommendations are statistically reliable.
|
||||
- **Attitude toward AI/ML:** mixed but practical; they like tools that reduce repetitive tasks and help them work faster, but they worry about complexity and blame if something goes wrong.
|
||||
- **Pain points AI should help with:** patient queue management, reminders, checklisting, basic decision support, documentation assistance, and training support.
|
||||
- **Hindrances in solution design:** low digital confidence in some sites, limited device availability, unstable connectivity, and insufficient time for training.
|
||||
- **Role in Dev/R&D:** frontline reality check; they expose whether the tool actually works in crowded, low-resource settings.
|
||||
- **How to approach them:** keep interfaces simple, mobile-friendly, low-click, and aligned with routine tasks; demonstrate value in under one minute.
|
||||
|
||||
#### User-Profile 4: Support Staff & Patients
|
||||
|
||||
- **Who they are:** vocational/basic-level workers, assistants, clerks, pharmacy support, and commune/district support staff.
|
||||
- **Age range:** about 18–30+ depending on role.
|
||||
- **Domain proficiency / responsibility:** foundational care support, dispensing, scheduling, registration, logistics, and basic patient flow handling.
|
||||
- **Working environment:** district/commune level, often with staff shortages, limited equipment, and very high dependency on manual coordination.
|
||||
- **What they know:** local workflow, patient flow, administrative routines, inventory, and practical daily operations.
|
||||
- **What they do not know:** clinical reasoning, advanced data interpretation, or how AI should be validated in medicine.
|
||||
- **Attitude toward AI/ML:** mostly utilitarian; they like anything that reduces repetitive admin work, but they fear systems that increase confusion or slow service.
|
||||
- **Pain points AI should help with:** registration, queue routing, reminder systems, inventory support, duplicate entry reduction, and scheduling.
|
||||
- **Hindrances in solution design:** low training time, little tolerance for complex setup, language/interface issues, and dependence on supervisor approval.
|
||||
- **Role in Dev/R&D:** workflow witnesses; they reveal hidden process bottlenecks and administrative failure points.
|
||||
- **How to approach them:** use concrete use cases, local language, visual flows, and very short onboarding; do not assume technical familiarity.
|
||||
- The User Profile (key-attribute) for these end-user in this product context includes (more focus on FRs)
|
||||
- [**→ The results was developed based on Gemini-DeepResearch given the methodology of Online Ethnography as follow:**](https://gemini.google.com/share/a4e65f2f47c8)
|
||||
|
||||
|
||||
| Platform Type | Selected Target Groups/Channels | Primary User Base Identified | Ethnographic Value & Observed Behaviors |
|
||||
| --- | --- | --- | --- |
|
||||
| **Professional Forums** | ykhoa.net, Vietnam Orthopaedic Association (VOA) portal | Radiologists, Surgeons, Medical Students, Rural Physicians | Discussions on uncompensated workloads, mHealth resistance in rural areas, formal professional networking, CME tracking. |
|
||||
| **Patient Forums** | Webtretho | Mothers, Adult Family Caregivers, Female Patients | Highlighting the primacy of family-centric healthcare decision-making, peer-to-peer sharing of alternative/folk remedies, high anxiety expression. |
|
||||
| **Facebook Groups** | Chẩn Đoán Hình Ảnh Online, Trung tâm Chấn thương chỉnh hình Tâm Anh, Thoát vị đĩa đệm | Radiologists, Orthopedic Surgeons, Patients | Clinicians sharing edge-case DICOMs for peer review. Patients posting raw X-rays seeking asynchronous online consultations. |
|
||||
| **YouTube Channels** | cdhaonline, USAC, ACC Chiropractic, Bác sĩ Trần Văn Phúc Official | Sonographers, Patients seeking conservative treatments | Visual tutorials for MSK ultrasound, patient consumption of non-pharmacological treatment videos (sciatica, knee pain). |
|
||||
|
||||
### **User Profile 5: The Diagnostic Radiologist (Bác sĩ Chẩn đoán hình ảnh)**
|
||||
|
||||
- **Who they are** within the MSK workflow is foundational; they are the primary data generators and the initial analytical layer. Diagnostic Radiologists in the Vietnamese public sector are highly specialized medical doctors tasked with interpreting complex, multi-modal imaging (X-ray, CT, MRI) and producing the definitive reports that guide all subsequent orthopedic or rheumatological interventions. They operate exactly at the critical intersection of deep human anatomy, pathological manifestation, and advanced digital imaging technology.
|
||||
- **Age range** for this demographic typically spans from 28 to 55 years old. This broad range encompasses highly tech-fluent, digitally native junior attendings who have trained entirely on digital screens, up to senior department heads who have had to actively transition their cognitive models from legacy analog film-based interpretation to modern digital PACS environments over the last critical decade.
|
||||
- **Domain proficiency / responsibility** for these users lies in rapid, highly accurate pattern recognition across various tissue densities and skeletal structures. They are explicitly responsible for detecting micro-fractures, subtle degenerative joint changes, early-stage bone tumors, and complex structural anomalies within the MSK system. Furthermore, in modernized environments like Bach Mai or Viet Duc, they are increasingly burdened with the responsibility of monitoring patient radiation dose exposure, a complex safety metric now deeply integrated into modern PACS software implementations.
|
||||
- **Working environment** `constraints are severe.` They predominantly work in high-stress, low-ambient-light reading rooms situated within severely overloaded central and provincial hospitals. Their work is highly asynchronous and entirely detached from direct, physical patient interaction. They face immense cognitive load due to the sheer, overwhelming volume of daily scans generated by a public healthcare system where patients often bypass primary care facilities to visit central hospitals directly, leading to massive backlogs.
|
||||
- **What they know** is exhaustive and deeply technical. They possess a profound understanding of skeletal anatomy, radiological physics, and pathological manifestations on high-resolution digital monitors. They know the intimate intricacies of complex DICOM viewers, including windowing, leveling, multi-planar reconstruction (MPR), and maximum intensity projections (MIP). They are intimately familiar with the limitations of specific imaging modalities and the technical artifacts that can obscure or mimic diagnoses.
|
||||
- **What they do not know** represents a critical clinical gap. They typically do not know the patient's full longitudinal clinical history or the nuanced, subjective experiences of the patient's localized pain, primarily because the clinical order notes provided by referring public sector physicians are often exceedingly brief due to extreme hospital overload. Furthermore, they are often unaware of the specific downstream surgical or conservative treatment decisions made by the orthopedic surgeons unless a formal multi-disciplinary case review is specifically initiated.
|
||||
- **Attitude toward AI/ML** among Vietnamese radiologists is a complex, evolving mixture of cautious optimism and defensive professional skepticism. The highly publicized and successful deployment of AI applications like VinBrain's DrAid for liver cancer screening and COVID-19 triage—which have won prestigious international awards and proven highly effective in real-world triage—has established a strong, tangible baseline of trust in algorithmic capabilities. They view AI highly favorably when it functions as an automated, invisible "second reader" that accurately highlights anomalies, quantifies tedious measurements, or triages obvious normal versus critical abnormal scans to help manage their overwhelming workload. However, profound resistance manifests if the AI is perceived as an opaque "black box" that disrupts established mental models without providing transparent, explainable reasoning.
|
||||
- **Pain points AI should help with** center entirely on cognitive fatigue and volume management. The primary pain point is the severe visual fatigue that leads to the risk of missing secondary, subtle findings (incidentalomas) due to the necessity of rapid patient throughput. AI must step in to automate highly tedious, repetitive measurements—such as joint space narrowing quantification in osteoarthritis or complex Cobb angles for scoliosis—and provide immediate professional objective checks. Furthermore, AI should actively assist in standardizing reports to reduce the heavy cognitive burden of continuous dictation and manual typing.
|
||||
- **Hindrances in solution design** are heavily rooted in behavioral economics and legacy system performance. A critical ethnographic hindrance observed in forums like ykhoa.net is the profound fear of uncompensated workload. If the proposed collaborative platform introduces features that require the radiologist to answer direct, text-based queries from patients or spend excessive time manually validating AI outputs, the system will face universal rejection in the public sector. Additionally, interface latency is a severe technical hindrance; any collaborative tool that slows down the rendering of massive, gigabyte-sized DICOM files will immediately disrupt their highly optimized, fast-paced diagnostic workflow.
|
||||
- **Role in Dev/R&D** processes is paramount; they serve as the ultimate ground-truth validators. In the R&D phase, their expertise is strictly required for accurately annotating MSK datasets, defining the specific clinical relevance of algorithmic edge cases, and rigorously evaluating the ergonomic efficiency of the DICOM viewing workspace. Their continuous feedback is absolutely paramount in finely tuning the AI's sensitivity-to-specificity ratio to prevent alert fatigue.
|
||||
- **How to approach & percieve them (as Product-Dev Team)** requires a strategy of deep professional respect and workflow preservation. The product development team must approach radiologists as overwhelmed, elite experts whose primary and most scarce currency is time. They must never be perceived as luddites or obstacles to digital innovation, but rather as the vital gatekeepers of clinical safety. Engagement and UX research should be focused entirely on invisible workflow optimization, strongly emphasizing how the platform will actively reduce their cognitive load and protect them from malpractice liabilities via AI-backed objective checks, rather than framing the platform as a disruptive tool meant to alter their fundamental diagnostic philosophy.
|
||||
|
||||
### **User Profile 7: The Rheumatologist & Orthopedic Surgeon (Bác sĩ Cơ xương khớp / Chấn thương chỉnh hình)**
|
||||
|
||||
- **Who they are** dictates the final clinical outcome. `These clinical specialists are the ultimate consumers of the diagnostic imaging data and the primary architects of the patient's comprehensive treatment plan.` Orthopedic surgeons focus heavily on mechanical and surgical interventions for acute trauma, severe degenerative diseases, and complex congenital anomalies, while rheumatologists manage chronic, systemic autoimmune conditions affecting the MSK system. They are the definitive authoritative figures in the patient's healthcare journey.
|
||||
- **Age range** is typically clustered between `30 and 60 years old.` Senior surgeons and veteran department heads hold immense, systemic influence over departmental procurement decisions and the formal adoption of new clinical pathways, while younger, highly digital attendings are significantly more likely to engage directly with novel digital patient consultation platforms and mobile health applications.
|
||||
- **Domain proficiency / responsibility** is vast and encompassing. They are highly proficient in complex clinical diagnosis, advanced surgical biomechanics, systemic pharmacology, and long-term, multi-variable disease management. Their massive responsibility encompasses `synthesizing fragmented patient histories`, `physical examination findings,` and varied, highly technical radiological data to `formulate a definitive, actionable treatment strategy.` Ultimately, they bear the `complete clinical, ethical, and legal responsibility for the patient's surgical or therapeutic outcomes.`
|
||||
- **Working environment** is notoriously highly fragmented and intensely pressured. Their daily environment is violently split between the sterile concentration of the operating theater, bustling and chaotic outpatient clinics, and crowded inpatient wards. In major public hospitals like Bach Mai, outpatient clinics are characterized by severe, systemic overcrowding, where a single senior doctor may be forced to see well over a hundred individual patients in a single grueling shift, necessitating incredibly rapid, highly decisive consultations that leave almost no time for extended dialogue.
|
||||
- **What they know** is highly pragmatic. They intimately `know the practical, actionable clinical implications` of specific radiological findings. They know precisely which structural abnormalities require immediate, aggressive surgical intervention, which can be managed conservatively with physical therapy, and the specific surgical approaches required. They possess a deep, practical understanding of patient psychology, recognizing that their patients are often highly anxious, vulnerable, `and highly susceptible to community misinformation.`
|
||||
- **What they do not know** highlights the necessity of the collaborative platform. They often lack the deep, pixel-level technical expertise of the dedicated radiologist regarding complex, multi-layered imaging artifacts or subtle MRI physics. `More significantly, and dangerously, they frequently do not know what the patient is actively doing outside the hospital walls—specifically, if the desperate patient is pursuing highly dangerous folk remedies, unregulated herbal treatments, or inappropriate, aggressive physical manipulations that could severely exacerbate their delicate condition.`
|
||||
- **Attitude toward AI/ML** is strictly pragmatic and highly outcome-oriented`. They are notably less interested in basic AI that merely identifies an obvious femur fracture (which they can easily see from across the room)` and are vastly more interested in complex AI that provides `deep predictive analytics.` For example, they **desire algorithms forecasting the exact rate of cartilage degeneration in knee osteoarthritis or** predicting the **statistical likelihood of hardware failure in spinal fusions based on localized bone density metrics**. Crucially, they view `AI as a massive potential force multiplier for patient education, provided it visually translates complex DICOM data into beautifully simple formats the patient can instantly understand, thereby saving precious consultation minutes.`
|
||||
- **Pain points AI should help with** are heavily centered around time deficits. The most critical pain point is the absolute `lack of time for comprehensive, empathetic patient education during severely overloaded clinical shifts`. AI `must explicitly help by automatically generating visually intuitive, patient-friendly, 3D summaries of the MSK issues directly from the complex DICOM workspace.` Furthermore, they desperately `need the platform to automatically aggregate highly fragmented patient data (historical X-rays, recent MRIs, scattered lab results) into a single, unified, rapid-consumption dashboard to exponentially expedite their clinical decision-making process.`
|
||||
- **Hindrances in solution design** revolve around communication boundaries. A major, system-killing hindrance is the risk of the platform inadvertently encouraging patients to continuously message the busy doctor for minor, non-clinical queries. The ethnographic data shows the fear of an uncompensated, unmanageable volume of digital communication is profound among Vietnamese clinicians. If the proposed collaborative workspace functions too much like a standard, open-ended messaging app (like Zalo or Messenger), it will be aggressively abandoned. The communication flow must be heavily structured and tightly gated.
|
||||
- **Role in Dev/R&D** requires strategic clinical oversight. They are absolutely crucial in defining the core clinical logic and the hierarchical, prioritized display of information within the platform's UI. They must strictly dictate what specific information is clinically relevant enough to be prominently displayed on the primary dashboard and what secondary data should be relegated to background menus. In R&D, they define the acceptable clinical thresholds for the AI's predictive models and dictate the required authoritative tone of the patient-facing educational modules.
|
||||
- **How to approach & percieve them (as Product-Dev Team)** requires positioning the product as a shield. The product development team must perceive these senior doctors as the entire system's core orchestrators who are operating under extreme, punishing time deficits. The UX approach should emphatically emphasize the platform's ability to act as a "clinical force multiplier." Design discussions should entirely center on how the interactive workspace can automate tedious patient education, definitively dispel dangerous, culturally prevalent folk medicine myths , and heavily streamline multi-disciplinary case reviews without adding a single minute or a single extra click to their currently exhausted workflow.
|
||||
|
||||
****
|
||||
|
||||
### **~~User Profile 6: The MSK Sonographer (Bác sĩ Siêu âm Cơ xương khớp)~~**
|
||||
|
||||
- **Who they are** within the clinical ecosystem represents a highly specialized, dynamic subset of diagnostics. MSK Sonographers in Vietnam are often distinct from general dark-room radiologists; they frequently operate as specialized clinicians or general practitioners who have undergone extensive, highly specific supplementary training to master the dynamic, real-time imaging of the musculoskeletal system. They physically operate the high-frequency ultrasound probes to intimately evaluate soft tissues, tendons, ligaments, and superficial bone structures in real-time motion.
|
||||
- **Age range** for this highly active profile is typically `25 to 50` years old. This group includes a significantly large cohort of younger, ambitious medical professionals actively seeking continuing medical education (CME) to differentiate their clinical skill sets, a trend heavily evidenced by the high demand and `rapid enrollment in structured online` and offline training courses offered by institutions like Pham Ngoc Thach University of Medicine and Medic Medical Center.
|
||||
- **Domain proficiency / responsibility** is uniquely tactile, highly kinetic, and heavily operator-dependent. Their primary proficiency lies in real-time functional assessments, such as directly observing tendon gliding or impingement during active patient movement. `They must possess a profound, innate spatial understanding to instantly translate two-dimensional, grainy ultrasound slices into comprehensive three-dimensional anatomical comprehension.` They are also frequently responsible for `precision-guiding interventional procedures, such as targeted joint injections or aspirations.`
|
||||
- **Working environment** constraints are physical and chaotic. Unlike radiologists isolated in quiet, dark reading rooms, sonographers work directly at the patient's bedside, in bustling emergency departments, or in highly trafficked dedicated ultrasound clinics. Their environment is inherently fast-paced, highly `interactive`, and requires continuous, complex physical maneuvering of both the bulky ultrasound equipment and the patient's body. They operate in a unique space where immediate, `continuous verbal communication with the patient regarding pain localization is standard practice`.
|
||||
- **What they know** is rooted in soft-tissue dynamics. They possess deep, specialized knowledge of soft-tissue pathologies, inflammatory markers visible via `Power Doppler,` and the dynamic, mechanical relationships between muscles, tendons, and joints. They explicitly know how to manually manipulate acoustic windows with the probe to actively bypass bone and gas artifacts. They intimately understand the immediate physical limitations and specific pain thresholds of the patient currently lying on their examination table.
|
||||
- **What they do not know** is the deeper structural context. They inherently lack the comprehensive, multi-system, deep-tissue overview provided by an MRI or CT scan. They cannot visualize deep bone marrow pathologies or assess the complete structural integrity of deep, complex joints (like the central hip) that are inaccessible to high-frequency ultrasound waves. Crucially, because ultrasound is so subjective, they are highly dependent on their own mechanical skill, meaning they often do not know how their real-time findings perfectly correlate with universally standardized baselines without subsequent external validation.
|
||||
- **Attitude toward AI/ML** is largely nascent but highly receptive to real-time, assistive technologies. Because MSK ultrasound is so heavily operator-dependent and challenging to master, there is a strong, articulated desire within the community for AI tools that can provide on-the-fly, augmented-reality style anatomical labeling, automate the tedious measurement of tendon thickness during dynamic movement, or standardize the grading of complex inflammatory signals (e.g., synovial hypertrophy in rheumatoid arthritis).
|
||||
- **Pain points AI should help with** revolve around standardization and documentation. The primary pain point is the extreme inter-operator variability inherent in ultrasound diagnostics. `AI must help by enforcing standardized image capture protocols`, providing real-time quality assurance of the acquired `images before the patient leaves the table, and offering automated comparative analysis against the patient's previous scans to track disease progression objectively.` Additionally, AI must desperately assist in generating automated, highly structured preliminary reports based directly on the captured images to `drastically reduce post-examination administrative typing time.`
|
||||
- **Hindrances in solution design** are heavily dominated by the physical constraints of the ultrasound examination itself. A sonographer's hands are perpetually occupied (one firmly manipulating the probe, the other adjusting the machine console), and their eyes are strictly locked on the ultrasound monitor. Therefore, any collaborative platform that requires extensive keyboard input, complex mouse navigation away from the primary viewing area, or disruptive screen toggling during the live examination will be physically unviable and immediately rejected in a clinical setting.
|
||||
- **Role in Dev/R&D** requires physical simulation. In the R&D process, MSK sonographers are absolutely critical for the live usability testing of the platform's interface within a physical, dynamic, bedside environment. They are essential for defining the strict parameters of real-time AI visual overlay features, rigorously ensuring that the added visual interface does not accidentally obscure critical, subtle diagnostic data on the monitor.
|
||||
- **How to approach & percieve them (as Product-Dev Team)** demands a focus on ergonomics. The product team must explicitly perceive sonographers as highly kinetic, physically engaged users. The UX approach must aggressively prioritize hands-free (e.g., `voice-activated or pedal-driven`) or highly streamlined, `single-tap interactions.` The team should focus collaborative discussions on how the new platform can act as an invisible, silent assistant that automatically captures, precisely labels, and seamlessly integrates their isolated ultrasound findings into the broader MSK collaborative workspace alongside the static X-rays and MRIs, effectively bridging the data gap between dynamic and static imaging modalities.
|
||||
|
||||
### [**User Profile 6: The Physical Therapist / PhysioTherapy (Chuyên Viên Vật Lý Trị Liệu)**](https://gemini.google.com/share/4399a5d9c85d)
|
||||
|
||||
- Who They Are & Age Range
|
||||
- **The Downstream Executor:** They are allied health professionals who operate strictly under physician-dictated prescriptions, meaning our system's Role-Based Access Control (RBAC) must legally prevent them from modifying primary medical diagnoses while giving them full autonomy over physical therapy application workflows.
|
||||
- **Highly Stratified Educational Tiers:** Approximately 53% to 54.9% of the user base holds entry-level vocational certificates or 3-year diplomas (Tier 1) with low clinical autonomy, while the remaining cohort consists of 4-year [B.Sc](http://b.sc/). or rare postgraduate practitioners (Tier 2) who handle advanced clinical reasoning. Less than 1% of the entire active field holds a Master's degree or higher.
|
||||
- **The Youth Avalanche Demographic:** This workforce is exceptionally young, with 36% to 40% falling into the 20 to 29 age bracket, and 41.7% between 30 and 39. Less than 10% to 14% of active practitioners are over the age of 40.
|
||||
- **Gender-Variable Physical Mechanics:** Females comprise 60% to 62% of the workforce in general urban hospital units, but the user base shifts to 78.9% male in high-intensity environments like military field hospitals. Your frontend design must feature scalable touch targets and layout adaptability to support varying hand sizes and physical environments.
|
||||
- **The Digital Native Vulnerability Paradox:** While their youth makes them highly tech-literate and fast adopters of cloud-based collaborative platforms, their lack of long-term clinical experience makes them highly prone to early-career physical burnout.
|
||||
- Domain Proficiency / Responsibility
|
||||
- **Hardware and Modality Calibration:** They are highly proficient in calibrating and operating physical treatment agents including TENS, NMES, Shortwave Diathermy, and therapeutic ultrasound machines.
|
||||
- **High-Exertion Manual Interventions:** They routinely execute physically taxing techniques such as joint mobilizations, deep tissue manipulation, trigger point releases, and intensive stretching protocols.
|
||||
- **Kinetic Patient Handling:** They are responsible for the heavy physical labor of lifting, transferring, and repositioning immobile, post-surgical, or neurological patients.
|
||||
- **Chaotic Context Switching:** Their workflow requires them to switch instantly between intense cognitive patient assessments, fine-tuning hardware configuration parameters, and performing exhausting physical manual therapy.
|
||||
- Working Environment
|
||||
- **The Resource-Constrained Public Sector:** The majority of your users operate in general hospitals characterized by extreme patient volumes, high ambient noise, and cramped treatment rooms often smaller than 20 square meters. Stationary desktop workstations are scarce, shared by dozens of staff members, and physically isolated in department corners.
|
||||
- **The State-of-the-Art Private Sector:** A smaller subset works in international or specialized sports clinics featuring advanced, combined therapy units, spacious layouts, and lower patient-to-therapist ratios that afford the time for deep data interaction.
|
||||
- **Austere and Remote Settings:** Some practitioners operate in military field installations or rural clinics with severe spatial constraints and highly volatile, low-bandwidth network infrastructure.
|
||||
- What They Know
|
||||
- **Granular Musculoskeletal Biomechanics:** They possess an expert, native understanding of human skeletal anatomy, muscle origins, insertions, innervations, and compound kinetic chains.
|
||||
- **Physiological Tissue Dynamics:** They understand exactly how different sound wave frequencies, heat agents, and electrical currents interact with and alter biological tissue elasticity.
|
||||
- **Acute Tactile Literacy:** They have highly conditioned palpatory skills, meaning they can instantly detect underlying structural changes, muscle guarding, and tissue density abnormalities purely through manual physical touch.
|
||||
- What They Do Not Know
|
||||
- **Raw Radiological Imaging Literacy:** Unless they hold specialized postgraduate certifications, reading raw, grayscale imaging data, recognizing subtle bone pathomechanics on an X-ray, or interpreting artifact anomalies on a scan is entirely outside their foundational education.
|
||||
- **Formal Evidence-Based Practice (EBP) Architecture:** The vast majority have received zero academic training in research methodologies. They do not know how to construct standardized PICO search strings or navigate academic index databases like Medline.
|
||||
- **AI Mathematical Frameworks:** Computational models, computer vision neural networks, and algorithmic prediction logic are completely opaque to them, meaning they perceive any unexplained AI output as an untrustworthy "black box".
|
||||
- **Integrated Longitudinal Records:** Due to systemic hospital data silos, they operate without any automated visibility into a patient's cross-departmental records, lab results, or concurrent medical prescriptions.
|
||||
- Attitude Toward AI/ML
|
||||
- **Defensiveness Against Clinical Devaluation:** Due to local market saturation and low entry-level compensation, younger therapists harbor economic anxiety about being commoditized. If our AI delivers authoritarian, rigid treatment mandates that bypass their personal clinical judgment, they will view the app as a threat and hostilely reject it.
|
||||
- **Appetite for Objective Validation Tools:** Conversely, they are highly receptive to features that provide empirical, quantifiable feedback (such as automatically calculating structural changes over a 6-week protocol) because it gives them undeniable data to validate their clinical expertise to skeptical doctors and patients.
|
||||
- **Acceptance Contingent on Transparancy:** They will only trust machine learning suggestions if the interface utilizes Explainable AI (XAI) principles, explicitly displaying the underlying clinical metrics behind a recommendation.
|
||||
- Pain Points AI Should Help With
|
||||
- **The Information Bottleneck:** Vietnamese Physiotherapists struggle to accurately target internal tissue pathologies during therapy because doctors rarely share full digital DICOM imaging data down the chain, providing only brief text prescription sheets, leading to semi-blind therapeutic execution via estimated surface anatomy and reduced treatment efficacy.
|
||||
- **The WMSD Crisis:** Vietnamese Physiotherapists struggle to maintain their own physical health and career longevity because constant manual strain and repetitive high-intensity clinical interventions overload their bodies, leading to an alarming 76.4% prevalence rate of painful, debilitating Work-Related Musculoskeletal Disorders.
|
||||
- **Crippling Time Poverty:** Vietnamese Physiotherapists struggle to implement evidence-based clinical practices and complete manual charting because extreme caseload pressures force them to treat 11 to 20+ patients per single daily shift, leading to severe operational time constraints, administrative exhaustion, and zero available time for typing medical records.
|
||||
- **`The Cross-Domain Literacy Gap`:** Vietnamese Physiotherapists struggle to accurately interpret diagnostic findings and seamlessly communicate with prescribing physicians because their training in kinetic biomechanics completely mismatches the physician’s world of raw radiological data, a barrier compounded by a massive 84% foreign language deficit and zero formal training in clinical statistics, leading to severe inter-departmental communication barriers, a reliance on subjective guesswork, and highly ineffective or contraindicated treatment suggestions.
|
||||
- Hindrances in Solution Design
|
||||
- **The Severe Foreign Language Barrier:** A massive 84% of Vietnamese PTs cite reading professional literature in foreign languages as their primary professional barrier, with only 25% reporting adequate English reading skills. Any unlocalized English UI elements or untranslated clinical documentation will result in total system failure.
|
||||
- **The "Informal Peer" Workflow Loop:** When facing a complex clinical decision, 96% rely on personal experience, 93% rely on informal chats with nearby peers, and 86% rely on old textbooks. They almost never utilize formal digital research dockets.
|
||||
- **Physical UI Contamination:** Because their hands are continuously covered in conductive ultrasound transmission gels, massage oils, or sweat, precise multi-touch menus and complex text inputs are completely unusable during patient sessions.
|
||||
- Role in Dev/R&D
|
||||
- **Physical Usability Grounding:** They must be utilized in ethnographic shadowing and beta testing to ensure our interface supports knuckle taps, gestures, or stylus interactions that function seamlessly when hands are covered in gel.
|
||||
- **FRONTEND CLINICAL VALIDATION:** They act as our real-time feedback loop, manually flagging when automated AI protocol recommendations conflict with a patient's physical pain tolerance or structural presentation.
|
||||
- How to Approach & Perceive Them (as Product-Dev Team)
|
||||
- **The "Clinical Athlete" Engineering Model:** Never design this software as if it is for a stationary desktop user. You must treat the PT as a highly kinetic, physically exhausted movement athlete who requires an interface optimized for extreme speed, low cognitive friction, and rapid physical access.
|
||||
- **The Ergonomic Value Proposition:** Frame the entire platform to the client not as a data tracking utility, but as an **intelligent digital exoskeleton** engineered specifically to absorb their administrative workload, protect their bodies from chronic injury, and visually elevate their clinical standing.
|
||||
|
||||
---
|
||||
|
||||
### Out-of-Domain Glossary (Physiotherapy & MSK Jargon)
|
||||
|
||||
### 1. Core Clinical Infrastructure & Data Types
|
||||
|
||||
- **DICOM (Digital Imaging and Communications in Medicine):** The international file format standard for medical imaging. A single DICOM file contains uncompressed high-fidelity image data bundled with heavy, embedded metadata tables detailing patient demographics, scanning coordinates, and precise machine calibration arrays.
|
||||
- **MSK (Musculoskeletal):** The complex anatomical framework comprising muscles, bones, cartilage, joints, ligaments, tendons, and bursa that support structure and enable physical locomotion.
|
||||
- **EBP (Evidence-Based Practice):** A medical decision-making framework requiring clinicians to systematically integrate the highest quality current published scientific research with their personal clinical expertise and the unique preferences of the patient.
|
||||
- **PICO Formulation:** A standardized syntax used to convert complex, vague clinical problems into clean, structured queries. It isolates variables into: **P**atient/Problem, **I**ntervention, **C**omparison, and **O**utcome.
|
||||
|
||||
### 2. Physical Therapy Modalities & Hardware Targets
|
||||
|
||||
- **Therapeutic Ultrasound:** A physical therapy device that emits high-frequency sound waves (0.8 to 3 MHz) directly into tissue via a handheld transducer to generate deep localized heat, dilate blood vessels, and increase collagen elasticity. Note: Unlike diagnostic ultrasound, it cannot generate an image on a screen.
|
||||
- **Diagnostic Ultrasound (Sonography):** An imaging modality used by doctors to capture and display real-time, dynamic grayscale representations of soft tissue boundaries, structural inflammation, and fluid accumulation.
|
||||
- **TENS (Transcutaneous Electrical Nerve Stimulation):** A modality that delivers low-voltage electrical currents via skin-surface electrodes to stimulate sensory nerves, effectively jamming the neural pathways that transmit pain signals to the brain.
|
||||
- **NMES (Neuromuscular Electrical Stimulation):** A device that passes electrical currents directly into a muscle belly to force involuntary muscle contractions, commonly utilized to reverse muscle atrophy or re-train pathways after neurological trauma.
|
||||
- **Shortwave Diathermy (SWD):** A high-frequency electromagnetic modality that uses deep wave currents to apply uniform therapeutic heat to extensive, deep-seated muscle masses and joint structures.
|
||||
- **Extracorporeal Shockwave Therapy:** A machine that emits high-energy acoustic shock waves into chronic, calcified soft tissues to intentionally cause localized micro-trauma, forcing the body to restart its natural healing and inflammatory cascades.
|
||||
|
||||
### 3. Manual Techniques & Structural Frameworks
|
||||
|
||||
- **Joint Mobilization:** A skilled, passive manual therapy technique where a PT applies targeted, graded physical forces at specific angles to glide, slide, or distract a patient's joint structures to restore normal structural motion.
|
||||
- **PNF (Proprioceptive Neuromuscular Facilitation):** An advanced therapeutic stretching framework that alternatingly triggers muscle contraction and passive relaxation against manual resistance to override standard neurological muscle-tightness reflexes.
|
||||
- **Palpatory Literacy:** The highly trained ability of a healthcare professional to identify underlying structural pathologies, knots, density variances, and fluid effusions solely via physical fingertip touch and manual assessment.
|
||||
- **RUSI (Rehabilitative Ultrasound Imaging):** The practice of using real-time ultrasound imaging during a physical therapy session as a visual biofeedback tool so both the therapist and the patient can immediately see if deep postural stabilization muscles are firing correctly.
|
||||
- **Anisotropic Artifact:** A physics-based visual error on an ultrasound scan where a completely healthy, uniform tendon appears falsely dark, hypoechoic, or "torn" simply because the acoustic beam did not strike the tissue fibers at a perfect 90-degree perpendicular angle.
|
||||
- **WMSDs (Work-Related Musculoskeletal Disorders):** Inflammation or structural damage to muscles, nerves, tendons, or ligaments that is directly caused, accelerated, or aggravated by repetitive work tasks, sustained poor ergonomics, or heavy physical lifting.
|
||||
|
||||
---
|
||||
|
||||
### **User Profile 8: The MSK Patient & Family Caregiver (Bệnh nhân Cơ xương khớp & Người nhà)**
|
||||
|
||||
- **Who they are** represents the most vulnerable and most populous user segment. This profile comprehensively encompasses individuals suffering from chronic or acute locomotive disorders, ranging from severe, mobility-limiting osteoarthritis to debilitating, acute herniated discs, alongside their immediate, highly involved family members. In the specific Vietnamese socio-cultural context, serious medical decisions are practically never made in isolation by the individual; adult children frequently and aggressively navigate the complex healthcare system on behalf of their elderly parents, while parents aggressively seek the absolute best care for their children, driven by deep, unwavering cultural imperatives.
|
||||
- **Age range** is bifurcated. The primary patients themselves are predominantly middle-aged to elderly (45 to 80+ years old), firmly representing the demographic most biologically afflicted by degenerative MSK diseases. However, the primary active users of the digital platform are very often their younger, more agile family caregivers (20 to 45 years old) who actually possess the required digital literacy to effectively navigate complex hospital apps, successfully book "green lane" digital appointments, and actively participate in digital social media health forums to crowdsource opinions.
|
||||
- **Domain proficiency / responsibility** regarding formal medical science is generally very low. However, they often possess high, albeit heavily skewed, experiential knowledge based on vast amounts of anecdotal evidence frantically gathered from community networks, massive online forums like Webtretho, or specialized, highly active Facebook groups. Their primary, exhausting responsibility is managing daily, chronic pain, attempting to adhere to complex treatment protocols (often poorly due to misunderstanding), and physically navigating the logistical nightmares of public hospital attendance.
|
||||
- **Working environment** is the home, the pharmacy, and the crowded hospital waiting room. Their environment sharply contrasts with the highly structured clinical setting. Their primary interaction with the healthcare system is deeply characterized by exceptionally long wait times, profound anxiety, and a pervasive feeling of disenfranchisement within the massive, impersonal machinery of central hospitals. They exist in a daily information ecosystem completely saturated with highly conflicting information regarding the efficacy of modern surgical medicine versus traditional, highly accessible remedies.
|
||||
- **What they know** is highly localized and experiential. They intimately know the exact nature of their own pain, their specific mobility limitations, and the severe financial strain their chronic condition places on the entire family unit. They explicitly know how to aggressively seek out alternative solutions when left frustrated by the extreme brevity of public hospital consultations, often turning in desperation to slick chiropractic videos on YouTube or utilizing unregulated folk healers operating within their local communities.
|
||||
- **What they do not know** poses a severe health risk. They fundamentally do not understand the underlying biomechanical realities of their specific conditions. They absolutely do not know how to read a standard X-ray or MRI, seeing only confusing, intimidating shades of gray. Crucially, they often do not understand the severe, irreversible risks associated with unscientific treatments, such as the distinct potential for permanent paralysis from incorrect, aggressive acupressure or massive spinal infection from applying raw leaves directly to spinal regions. They also deeply struggle to comprehend the realistic limitations of surgical interventions, often erroneously expecting immediate, permanent, pain-free cures.
|
||||
- **Attitude toward AI/ML** is largely unformed, highly malleable, but exceptionally susceptible to the precise UI framing of the technology. If the AI is expertly presented as a high-tech, highly objective authority that visually validates their human doctor's hurried diagnosis, it can significantly and immediately increase institutional trust. Because ethnographic data proves many patients already actively seek second opinions on Facebook groups by desperately posting their raw scans , an AI tool integrated into a patient portal that provides an immediate, highly understandable, beautifully rendered visual analysis of their DICOM images would be viewed as highly empowering, deeply reassuring, and vastly superior to social media crowdsourcing.
|
||||
- **Pain points AI should help with** are driven by fear of the unknown. The overwhelming pain point is profound confusion and anxiety stemming directly from a lack of comprehensible, visually accessible information. AI must explicitly help by transforming terrifying, complex DICOM scans into clear, color-coded, interactive 3D anatomical models that explicitly illustrate their specific injury or degeneration in laymen's terms. The platform must actively provide localized, deeply culturally contextualized educational content that patiently explains *why* specific medical treatments are necessary and explicitly *why* specific, popular folk remedies like leaf-wrapping are actively dangerous.
|
||||
- **Hindrances in solution design** are heavily rooted in digital inequity. The digital divide is the most profound, systemic hindrance. Elderly patients exhibit significant, stubborn resistance and a literal inability to adopt new mobile technologies, often actively preferring traditional, highly inefficient "arrive early and wait" methods over interacting with modern digital scheduling apps, directly undermining efficiency efforts like the "luồng xanh". Furthermore, if the platform's educational material is heavily text-based or overly academic in tone, it will be immediately ignored in favor of highly engaging, visually stimulating, but medically inaccurate, social media video content. The necessary reliance on a proxy user (the younger caregiver) vastly complicates the technical design of direct-to-patient privacy, data security, and consent architectures.
|
||||
- **Role in Dev/R&D** demands emotional intelligence from the researchers. Patients and their caregivers must be central to the rigorous evaluation of the platform's accessibility, emotional resonance, and comprehensibility. Their required role in R&D involves deep usability testing of the patient portal, specifically ensuring that the AI-translated educational visuals are actually lowering clinical anxiety rather than inadvertently increasing it by displaying terrifyingly realistic pathology. They provide critical, irreplaceable feedback on the emotional tone of the entire user interface.
|
||||
- **How to approach & percieve them (as Product-Dev Team)** requires profound empathy and accessible design principles. The product development team must perceive this vast profile with deep empathy, explicitly recognizing them as highly vulnerable users frantically navigating a complex, high-anxiety ecosystem. The UX approach must aggressively prioritize extreme, uncompromising simplicity, absolute visual clarity, and highly robust accessibility features (e.g., massive font scaling, high-contrast modes, voice-over capabilities). The team must strategically view the younger caregiver as the primary digital conduit to the elderly patient, thereby requiring the design of sophisticated dual-access interface architectures that allow entire families to collaboratively view, discuss, and understand the educational outputs securely generated from the clinical workspace.
|
||||
- The Multi-hat Scenarios (where 1 doctors shall involve multiple-hats of professional is usual in Vietnam)
|
||||
|
||||
---
|
||||
|
||||
Reference in New Issue
Block a user