update readme

Add information about external validation

README.md

@@ -14,9 +14,9 @@ pipeline_tag: image-classification
 # Agentic Thyroid ResNet-18 — Ultrasound Nodule Malignancy Classifier
 
 > ⚠️ **RESEARCH USE ONLY — NOT FOR CLINICAL USE.** This model is a research
-> artifact trained on a single retrospective dataset. It has **not** been
-> externally validated and **must not** be used for diagnosis, screening, or any
-> clinical decision-making. External, prospective validation is required before
+> artifact trained on a single retrospective dataset and is validated only one external dataset. It
+> **must not** be used for diagnosis, screening, or any
+> clinical decision-making. Multiple external, prospective validation is required before
 > any clinical consideration.
 
 A ResNet-18 binary classifier that predicts the probability that a cropped
@@ -155,7 +155,7 @@ python evaluate_external.py \
 
 ## Limitations, bias, and leakage concerns
 
-- **Single-source dataset; no external validation.** Performance on data from
+- **Single-source dataset** Performance on data from
   other scanners, institutions, or populations is unknown and likely lower.
 - **Cropped-ROI inputs.** The model expects nodule-cropped images like TN5000;
   whole-frame ultrasound will be out of distribution.
@@ -169,9 +169,11 @@ python evaluate_external.py \
 - **Leakage checks were exhaustive within the available signal** (exact pixel
   hashing + filename-ID overlap, all zero) but cannot rule out near-duplicate or
   same-patient-different-image leakage if such structure exists in the source.
+  But the original article describing the dataset states that "Only one representative image among the images of the same perspective of the patient is reserved, to prevent the risk of data leakage and ensure the dataset can be partitioned at the patient level;"
 
 ## ⚠️ External validation required before clinical use
 
+At present this model has only one external validation.
 This model **requires independent, ideally prospective, multi-site external
 validation** before any clinical consideration. It is released for research
 reproducibility only.