${esc(x)}

`).join(\"\");\n\n return `Thyroid Report\n\n\n

Thyroid Disease Detection Report

Generated ${esc(date)} · AI Clinical Decision Support

\n\n

Patient Details

Age: ${esc(p.age)}

Gender: ${esc(p.gender)}

Heart Rate: ${esc(p.heartRate)} bpm

Pregnant: ${p.pregnant ? \"Yes\" : \"No\"}

Family History: ${p.familyHistory ? \"Yes\" : \"No\"}

Fatigue: ${p.fatigue ? \"Yes\" : \"No\"}

\n\n

Laboratory Values

TSH: ${esc(p.tsh)} mIU/L

T3: ${esc(p.t3)} ng/dL

TT4: ${esc(p.tt4)} µg/dL

T4U: ${esc(p.t4u)}

FTI: ${esc(p.fti)}

\n\n

Prediction

Diagnosis: ${esc(r.diagnosis)}\n Confidence: ${esc(r.confidence)}%

Risk Level: ${esc(r.risk)}

${esc(r.explanation)}

\n\n

Contributing Factors (Explainable AI)

\n ${rows}

Feature	Finding	Direction

\n\n

Health Recommendations

${recs}

\n\n

⚠️ This automated report is decision support only and does not replace a licensed physician's diagnosis.

\n`;\n}\n","deleted","label",[124],"13c2874a-5c31-11f1-9fa1-ee650bb23af1","predict.ts","13c287f4-5c31-11f1-9fa1-ee650bb23af1",1,["D",1780150127979],"// Thyroid disease detection engine.\n// A transparent clinical scoring model that mimics ML output with explainable\n// per-feature contributions (SHAP-style). Reference ranges are standard adult labs.\n\nexport interface PatientInput {\n age: number;\n gender: \"male\" | \"female\";\n tsh: number; // mIU/L (ref 0.4 - 4.0)\n t3: number; // ng/dL (ref 80 - 200)\n tt4: number; // µg/dL (ref 5.0 - 12.0)\n t4u: number; // ratio (ref 0.7 - 1.8)\n fti: number; // index (ref 6 - 12)\n familyHistory: boolean;\n fatigue: boolean;\n weightChange: \"gain\" | \"loss\" | \"none\";\n heartRate: number; // bpm\n pregnant: boolean;\n}\n\nexport type Diagnosis = \"Hypothyroidism\" | \"Hyperthyroidism\" | \"Normal\";\nexport type Risk = \"Low Risk\" | \"Medium Risk\" | \"High Risk\";\n\nexport interface Contribution {\n feature: string;\n detail: string;\n weight: number; // signed: + pushes hypo, - pushes hyper\n}\n\nexport interface PredictionResult {\n diagnosis: Diagnosis;\n confidence: number; // 0-100\n risk: Risk;\n scores: { hypo: number; hyper: number; normal: number };\n contributions: Contribution[];\n explanation: string;\n recommendations: string[];\n}\n\nconst REF = {\n tsh: [0.4, 4.0],\n t3: [80, 200],\n tt4: [5.0, 12.0],\n t4u: [0.7, 1.8],\n fti: [6, 12],\n};\n\nexport function predict(p: PatientInput): PredictionResult {\n const contributions: Contribution[] = [];\n let hypo = 0;\n let hyper = 0;\n\n const add = (feature: string, detail: string, weight: number) => {\n contributions.push({ feature, detail, weight });\n if (weight > 0) hypo += weight;\n else hyper += -weight;\n };\n\n // TSH — the single most important marker (inverse relationship with thyroid activity)\n if (p.tsh > REF.tsh[1]) {\n const w = Math.min(40, 18 + (p.tsh - REF.tsh[1]) * 4);\n add(\"TSH\", `High TSH (${p.tsh} mIU/L)`, w);\n } else if (p.tsh < REF.tsh[0]) {\n const w = Math.min(40, 18 + (REF.tsh[0] - p.tsh) * 30);\n add(\"TSH\", `Low TSH (${p.tsh} mIU/L)`, -w);\n } else {\n add(\"TSH\", `Normal TSH (${p.tsh} mIU/L)`, 0);\n }\n\n // T3\n if (p.t3 < REF.t3[0]) add(\"T3\", `Low T3 (${p.t3} ng/dL)`, 15);\n else if (p.t3 > REF.t3[1]) add(\"T3\", `High T3 (${p.t3} ng/dL)`, -18);\n else add(\"T3\", `Normal T3 (${p.t3} ng/dL)`, 0);\n\n // TT4\n if (p.tt4 < REF.tt4[0]) add(\"TT4\", `Low TT4 (${p.tt4} µg/dL)`, 14);\n else if (p.tt4 > REF.tt4[1]) add(\"TT4\", `High TT4 (${p.tt4} µg/dL)`, -16);\n else add(\"TT4\", `Normal TT4 (${p.tt4} µg/dL)`, 0);\n\n // FTI\n if (p.fti < REF.fti[0]) add(\"FTI\", `Low FTI (${p.fti})`, 10);\n else if (p.fti > REF.fti[1]) add(\"FTI\", `High FTI (${p.fti})`, -12);\n else add(\"FTI\", `Normal FTI (${p.fti})`, 0);\n\n // T4U (mild signal)\n if (p.t4u < REF.t4u[0]) add(\"T4U\", `Low T4U (${p.t4u})`, 4);\n else if (p.t4u > REF.t4u[1]) add(\"T4U\", `High T4U (${p.t4u})`, -4);\n\n // Heart rate — clinical symptom\n if (p.heartRate > 100) add(\"Heart Rate\", `Tachycardia (${p.heartRate} bpm)`, -10);\n else if (p.heartRate < 60) add(\"Heart Rate\", `Bradycardia (${p.heartRate} bpm)`, 8);\n\n // Weight change\n if (p.weightChange === \"gain\") add(\"Weight\", \"Unexplained weight gain\", 8);\n else if (p.weightChange === \"loss\") add(\"Weight\", \"Unexplained weight loss\", -8);\n\n // Fatigue (slight hypo lean)\n if (p.fatigue) add(\"Fatigue\", \"Reports fatigue\", 5);\n\n // Family history raises overall suspicion equally\n if (p.familyHistory) {\n const lean = hypo >= hyper ? 4 : -4;\n add(\"Family History\", \"Positive family history\", lean);\n }\n\n const total = hypo + hyper;\n let diagnosis: Diagnosis;\n let confidence: number;\n const normalScore = Math.max(0, 40 - total);\n\n const sum = hypo + hyper + normalScore || 1;\n const scores = {\n hypo: Math.round((hypo / sum) * 100),\n hyper: Math.round((hyper / sum) * 100),\n normal: Math.round((normalScore / sum) * 100),\n };\n\n if (hypo >= hyper && hypo > 20) {\n diagnosis = \"Hypothyroidism\";\n confidence = scores.hypo;\n } else if (hyper > hypo && hyper > 20) {\n diagnosis = \"Hyperthyroidism\";\n confidence = scores.hyper;\n } else {\n diagnosis = \"Normal\";\n confidence = Math.max(scores.normal, 60);\n }\n\n // Risk level\n let risk: Risk = \"Low Risk\";\n if (diagnosis !== \"Normal\") {\n if (total > 55) risk = \"High Risk\";\n else if (total > 30) risk = \"Medium Risk\";\n else risk = \"Low Risk\";\n }\n\n // Explanation — top 2 drivers\n const drivers = [...contributions]\n .filter((c) => c.weight !== 0)\n .sort((a, b) => Math.abs(b.weight) - Math.abs(a.weight))\n .slice(0, 2);\n let explanation: string;\n if (diagnosis === \"Normal\") {\n explanation = \"All key thyroid markers fall within normal reference ranges.\";\n } else {\n const pct = Math.round(\n (drivers.reduce((s, d) => s + Math.abs(d.weight), 0) / (total || 1)) * 100,\n );\n explanation = `${drivers.map((d) => d.detail).join(\" and \")} contributed ${pct}% to the prediction of ${diagnosis}.`;\n }\n\n return {\n diagnosis,\n confidence,\n risk,\n scores,\n contributions: contributions.filter((c) => c.weight !== 0),\n explanation,\n recommendations: recommend(diagnosis, risk, p),\n };\n}\n\nfunction recommend(d: Diagnosis, risk: Risk, p: PatientInput): string[] {\n const recs: string[] = [];\n if (d === \"Hypothyroidism\") {\n recs.push(\"🩺 Consult an endocrinologist about levothyroxine therapy.\");\n recs.push(\"🥗 Diet: iodine-rich foods, selenium (Brazil nuts), avoid excess soy & raw cruciferous veg.\");\n recs.push(\"🏃 Exercise: regular low-impact cardio to support metabolism.\");\n recs.push(\"🔁 Re-test TSH/T4 in 6–8 weeks to monitor response.\");\n } else if (d === \"Hyperthyroidism\") {\n recs.push(\"🩺 Urgent endocrinology referral; discuss anti-thyroid medication.\");\n recs.push(\"🥗 Diet: adequate calories & calcium; limit iodine and caffeine.\");\n recs.push(\"🧘 Manage stress; avoid stimulants that worsen palpitations.\");\n recs.push(\"❤️ Monitor heart rate and blood pressure regularly.\");\n } else {\n recs.push(\"✅ Thyroid function appears normal — maintain a balanced diet and routine checkups.\");\n recs.push(\"🔁 Re-screen annually, or sooner if symptoms develop.\");\n }\n if (p.pregnant) recs.push(\"🤰 Pregnancy detected: thyroid levels need close monitoring — inform your OB-GYN.\");\n if (risk === \"High Risk\") recs.push(\"⚠️ High risk: seek medical consultation promptly.\");\n recs.push(\"ℹ️ This tool is decision support only and does not replace professional diagnosis.\");\n return recs;\n}\n",[141],"198f8678-5c31-11f1-a7e3-ee650bb23af1","main.tsx",2,["D",1780150137711],"http","/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport { Hono } from \"npm:hono@4.4.0\";\nimport { predict, type PatientInput } from \"./predict.ts\";\n\nconst app = new Hono();\n\n// Serve the front-end HTML page (static asset, no logic in template strings).\napp.get(\"/\", async (c) => {\n const { readFile } = await import(\"https://esm.town/v/std/utils/index.ts\");\n const html = await readFile(\"/index.html\", import.meta.url);\n return c.html(html);\n});\n\n// Prediction API\napp.post(\"/api/predict\", async (c) => {\n try {\n const body = (await c.req.json()) as PatientInput;\n const result = predict(normalize(body));\n return c.json(result);\n } catch (e) {\n return c.json({ error: String(e) }, 400);\n }\n});\n\n// Plain-text printable medical report (browser \"Save as PDF\")\napp.post(\"/api/report\", async (c) => {\n const body = (await c.req.json()) as PatientInput;\n const r = predict(normalize(body));\n const { buildReportHtml } = await import(\"./report.ts\");\n return c.html(buildReportHtml(body, r));\n});\n\nfunction normalize(b: any): PatientInput {\n return {\n age: Number(b.age) || 0,\n gender: b.gender === \"male\" ? \"male\" : \"female\",\n tsh: Number(b.tsh) || 0,\n t3: Number(b.t3) || 0,\n tt4: Number(b.tt4) || 0,\n t4u: Number(b.t4u) || 0,\n fti: Number(b.fti) || 0,\n familyHistory: !!b.familyHistory,\n fatigue: !!b.fatigue,\n weightChange: [\"gain\", \"loss\", \"none\"].includes(b.weightChange) ? b.weightChange : \"none\",\n heartRate: Number(b.heartRate) || 0,\n pregnant: !!b.pregnant,\n };\n}\n\nexport default app.fetch;\n",[148],"2c9b67f0-5c31-11f1-9885-ee650bb23af1","index.html","2c9b6890-5c31-11f1-9885-ee650bb23af1",4,["D",1780150169664],"\n\n\n \n \n Thyroid Disease Detection\n \n \n\n\n

🦋 Thyroid Disease Detection

AI-powered detection, risk scoring & explainable insights

\n\n \n\n

\n\n \n

\n View source\n

\n\n\n",[155],"391a34de-5c31-11f1-9b8c-ee650bb23af1","app.js","391a359c-5c31-11f1-9b8c-ee650bb23af1",5,["D",1780150190618],"// Client-side logic for the thyroid detection UI.\nconst form = document.getElementById(\"form\");\nconst result = document.getElementById(\"result\");\ndocument.getElementById(\"vt\").href =\n \"https://www.val.town/x/vaishnavipm/thyroid-detection\";\n\nfunction collect() {\n const fd = new FormData(form);\n return {\n age: fd.get(\"age\"), gender: fd.get(\"gender\"),\n tsh: fd.get(\"tsh\"), t3: fd.get(\"t3\"), tt4: fd.get(\"tt4\"),\n t4u: fd.get(\"t4u\"), fti: fd.get(\"fti\"), heartRate: fd.get(\"heartRate\"),\n weightChange: fd.get(\"weightChange\"),\n familyHistory: fd.get(\"familyHistory\") === \"on\",\n fatigue: fd.get(\"fatigue\") === \"on\",\n pregnant: fd.get(\"pregnant\") === \"on\",\n };\n}\n\nconst diagColor = {\n Hypothyroidism: \"bg-amber-100 text-amber-800\",\n Hyperthyroidism: \"bg-rose-100 text-rose-800\",\n Normal: \"bg-emerald-100 text-emerald-800\",\n};\nconst riskColor = {\n \"Low Risk\": \"bg-emerald-500\", \"Medium Risk\": \"bg-amber-500\", \"High Risk\": \"bg-rose-500\",\n};\n\nform.addEventListener(\"submit\", async (e) => {\n e.preventDefault();\n const payload = collect();\n const res = await fetch(\"/api/predict\", {\n method: \"POST\", headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify(payload),\n });\n const r = await res.json();\n render(r, payload);\n});\n\nfunction bar(label, pct, color) {\n return `

${label}${pct}%

`;\n}\n\nfunction render(r, payload) {\n const contribs = r.contributions\n .sort((a, b) => Math.abs(b.weight) - Math.abs(a.weight))\n .map((c) => {\n const dir = c.weight > 0 ? \"Hypo\" : \"Hyper\";\n const w = Math.min(100, Math.abs(c.weight) * 2.2);\n const col = c.weight > 0 ? \"bg-amber-400\" : \"bg-rose-400\";\n return `

\n ${c.feature} — ${c.detail}→ ${dir}\n

`;\n }).join(\"\");\n\n result.innerHTML = `\n

Predicted Diagnosis

\n ${r.diagnosis}\n

Confidence

${r.confidence}%

Risk Level

\n ${r.risk}\n

💡 ${r.explanation}

\n\n

Class Probabilities

\n ${bar(\"Hypothyroidism\", r.scores.hypo, \"bg-amber-500\")}\n ${bar(\"Hyperthyroidism\", r.scores.hyper, \"bg-rose-500\")}\n ${bar(\"Normal\", r.scores.normal, \"bg-emerald-500\")}\n

\n\n

🧠 Explainable AI — Contributing Factors

\n ${contribs || '

All markers normal.

'}\n

\n\n

📋 Personalized Recommendations

${x}

\n\n `;\n\n result.classList.remove(\"hidden\");\n result.scrollIntoView({ behavior: \"smooth\" });\n\n document.getElementById(\"pdf\").addEventListener(\"click\", async () => {\n const res = await fetch(\"/api/report\", {\n method: \"POST\", headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify(payload),\n });\n const html = await res.text();\n const w = window.open(\"\", \"_blank\");\n w.document.write(html);\n w.document.close();\n });\n}\n",[162],"4a07b294-5c31-11f1-9885-ee650bb23af1","README.md","4a07b384-5c31-11f1-9885-ee650bb23af1",["D",1780150219029],"# 🦋 Thyroid Disease Detection\n\nAn AI-powered clinical decision-support web app that predicts **Hypothyroidism**,\n**Hyperthyroidism**, or **Normal** thyroid function from patient lab + clinical data,\nexplains *why* (SHAP-style feature contributions), scores risk, and generates a\nprintable medical report.\n\n> ⚠️ Decision support only — not a substitute for a licensed physician's diagnosis.\n\n## Features\n\n- **Multi-class prediction** with confidence + per-class probabilities\n- **Risk scoring**: Low / Medium / High\n- **Explainable AI** — ranked contributing factors (e.g. *\"High TSH and Low T3 contributed 75% to Hypothyroidism\"*)\n- **Personalized recommendations** (diet, exercise, medical advice, pregnancy alerts)\n- **Printable PDF report** (via browser print)\n\n## How it works\n\n```mermaid\nflowchart LR\n A[Patient form] -->|POST /api/predict| B[predict.ts engine]\n B --> C{Score TSH, T3,
TT4, T4U, FTI,
symptoms}\n C --> D[Diagnosis + Confidence]\n C --> E[Risk Level]\n C --> F[Feature Contributions XAI]\n D & E & F --> G[UI result cards]\n G -->|POST /api/report| H[report.ts → printable PDF]\n```\n\nThe engine in `predict.ts` is a **transparent clinical scoring model** using standard\nadult reference ranges (TSH 0.4–4.0 mIU/L, etc.). High TSH → hypothyroid; low TSH →\nhyperthyroid, weighted alongside T3/TT4/FTI and symptoms (heart rate, weight change,\nfatigue, family history). Every score is traceable, which is what powers the\nexplainability view — no black box.\n\n## Files\n\n| File | Purpose |\n|------|---------|\n| `main.tsx` | Hono HTTP server + API routes |\n| `predict.ts` | Prediction + risk + explainability engine |\n| `report.ts` | Printable HTML/PDF report builder |\n| `index.html` | Patient input UI (Twind styling) |\n| `app.js` | Client-side form + results rendering |\n\n## Extending to a trained ML model\n\nTo swap the rule-based engine for a real trained model (Random Forest / XGBoost /\nneural net per the project brief), keep the same `predict()` signature and:\n1. Train offline on a labelled thyroid dataset (e.g. UCI Thyroid).\n2. Export the model (ONNX / JSON weights) and load it in `predict.ts`.\n3. Use SHAP values at training time to seed the contribution weights.\n\nThe UI and report layers need no changes.\n",[169],"pulls","forks","envVars",{"_180":54,"_182":54},"connections",{"_180":54,"_181":54},"count","unfilledCount","emptyCount",{"_20":168,"_13":169,"_125":170,"_127":-5,"_128":21,"_129":30,"_131":50,"_22":205,"_29":54,"_133":103,"_135":172,"_137":38,"_138":-5,"_190":206,"_91":207},{"_20":161,"_13":162,"_125":163,"_127":-5,"_128":21,"_129":164,"_131":50,"_22":202,"_29":54,"_133":134,"_135":166,"_137":38,"_138":-5,"_190":203,"_91":204},{"_20":154,"_13":155,"_125":156,"_127":-5,"_128":21,"_129":157,"_131":50,"_22":199,"_29":54,"_133":103,"_135":159,"_137":38,"_138":-5,"_190":200,"_91":201},{"_20":147,"_13":148,"_125":48,"_127":-5,"_128":21,"_129":149,"_131":50,"_22":196,"_29":54,"_133":151,"_135":152,"_137":38,"_138":-5,"_190":197,"_91":198},{"_20":140,"_13":141,"_125":142,"_127":-5,"_128":21,"_129":143,"_131":50,"_22":193,"_29":54,"_133":134,"_135":145,"_137":38,"_138":-5,"_190":194,"_91":195},{"_20":123,"_13":124,"_125":126,"_127":-5,"_128":21,"_129":130,"_131":50,"_22":189,"_29":54,"_133":134,"_135":136,"_137":38,"_138":-5,"_190":191,"_91":192},["D",1780150154706],"children",[],[124],["D",1780150127979],[],[141],["D",1780150137711],[],[148],["D",1780150169664],[],[155],["D",1780150190618],[],[162],["D",1780150219029],[],[169],["D",1780150092882],{"_11":12,"_89":85},{"_11":12,"_82":83,"_84":85,"_45":46,"_86":87,"_88":38},{"_29":30,"_13":26,"_20":21,"_27":28,"_31":-5,"_32":-5,"_77":-5,"_78":-5,"_79":-5,"_80":14,"_81":12},"ctx",{"_70":-5,"_111":245,"_246":247,"_248":249,"_250":38,"_251":38,"_252":38,"_253":50,"_254":-5,"_255":-7,"_256":-7,"_257":54,"_258":38,"_259":38,"_260":-5,"_261":38,"_262":38,"_263":38,"_264":-5},"ENV",{"_216":217,"_218":219,"_220":26,"_221":222,"_223":224,"_225":226,"_227":228,"_229":230,"_231":232,"_233":234,"_235":236,"_237":238,"_239":240,"_241":242,"_243":244},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","a3e067d4463eea85f6334f8e583e8516ae58c4f6","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Branches1

Branches
1