init

Medical Diagnosis App/.streamlit/config.toml

@@ -0,0 +1,6 @@
+[theme]
+primaryColor="#38b2ac"
+backgroundColor="#1a202c"
+secondaryBackgroundColor="#2d3748"
+textColor="#e2e8f0"
+font='monospace'

Medical Diagnosis App/.vscode/settings.json

@@ -0,0 +1,4 @@
+{
+    "editor.tabCompletion": "on",
+    "diffEditor.codeLens": true
+}

Medical Diagnosis App/Home.py

@@ -0,0 +1,25 @@
+import streamlit as st
+from PIL import Image
+
+
+# Add a title
+st.set_page_config(page_title="Select Diagnosis", layout="centered")
+
+
+
+st.title("Medical Diagnosis App")
+
+st.markdown("")
+st.markdown("<li> Currently Brain Tumors , Xrays and Skin Leison Analysis are ready for diagnosis </li>"
+            "<li>The Models also explain what area in the images is the cause of diagnosis </li>"
+            "<li>Currently the models are trained on a small dataset and will be trained on a larger dataset in the future</li>"
+            '<li> The Application also provides generated information on how to diagnose the disease and what should the patient do in that case</li>'
+            ,unsafe_allow_html=True)
+
+with st.sidebar.container():
+    image = Image.open("/Users/vikram/Downloads/Meditechlogo.png")
+    st.image(image, caption='Meditech',use_column_width=True)
+
+
+
+

Medical Diagnosis App/brain_labels.json

@@ -0,0 +1,6 @@
+{
+    "0":"Glinomia",
+    "1": "Meningomia",
+    "2":"notumar",
+    "3": "pituary"
+}

Medical Diagnosis App/labels.json

@@ -0,0 +1,6 @@
+{
+    "0":"Covid19",
+    "1": "Normal",
+    "2":"Pneumonia",
+    "3": "Tuberculosis"
+}

Medical Diagnosis App/models/brain_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52a7a2393337e2dedca977e4729df96b2f55579c58cbd2263922d0f8752fd866
+size 79728020

Medical Diagnosis App/models/eye_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e83d4f41fd93bd35208cb9a1643158ee33b781ea111105fe614e2738edb3fe7
+size 27238762

Medical Diagnosis App/models/timm_skin_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5163d00507a13ccd735162fd43dd35b16cbb901bc06d407a4deb1cef194a0e2
+size 16408803

Medical Diagnosis App/models/timm_xray_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe8a9cda4e25a731216bb6503f36757e9ebde7251f2c1f34c1aa3489707419c4
+size 93909024

Medical Diagnosis App/pages/Brain.py

@@ -0,0 +1,156 @@
+import streamlit as st
+from PIL import Image
+import torch.nn as nn
+import timm
+import torch
+import torchmetrics
+from torchmetrics import F1Score,Recall,Accuracy
+import torch.optim.lr_scheduler as lr_scheduler
+import torchvision.models as models
+import lightning.pytorch as pl
+import torchvision
+from lightning.pytorch.loggers import WandbLogger
+import shap
+import matplotlib.pyplot as plt
+import json 
+from transformers import pipeline, set_seed
+from transformers import BioGptTokenizer, BioGptForCausalLM
+text_model = BioGptForCausalLM.from_pretrained("microsoft/biogpt")
+tokenizer = BioGptTokenizer.from_pretrained("microsoft/biogpt")
+labels_path = '/Users/vikram/Python/Medical Diagnosis App/brain_labels.json'
+from captum.attr import DeepLift , visualization
+
+with open(labels_path) as json_data:
+    idx_to_labels = json.load(json_data)
+
+
+
+class FineTuneModel(pl.LightningModule):
+    def __init__(self, model_name, num_classes, learning_rate, dropout_rate,beta1,beta2,eps):
+        super().__init__()
+        self.model_name = model_name
+        self.num_classes = num_classes
+        self.learning_rate = learning_rate
+        self.beta1 = beta1
+        self.beta2 = beta2
+        self.eps = eps
+        self.dropout_rate = dropout_rate
+        self.model = timm.create_model(self.model_name, pretrained=True,num_classes=self.num_classes)
+        self.loss_fn = nn.CrossEntropyLoss()
+        self.f1 = F1Score(task='multiclass', num_classes=self.num_classes)
+        self.recall = Recall(task='multiclass', num_classes=self.num_classes)
+        self.accuracy = Accuracy(task='multiclass', num_classes=self.num_classes)
+        
+        #for param in self.model.parameters():
+            #param.requires_grad = True
+        #self.model.classifier= nn.Sequential(nn.Dropout(p=self.dropout_rate),nn.Linear(self.model.classifier.in_features, self.num_classes))
+        #self.model.classifier.requires_grad = True
+            
+
+    def forward(self, x):
+        return self.model(x)
+
+    def training_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('train_loss', loss,on_step=False,on_epoch=True)
+        self.log('train_acc', acc,on_step=False,on_epoch = True)
+        self.log('train_f1',f1,on_step=False,on_epoch=True)
+        self.log('train_recall',recall,on_step=False,on_epoch=True)
+        return loss
+            
+    def validation_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('val_loss', loss,on_step=False,on_epoch=True)
+        self.log('val_acc', acc,on_step=False,on_epoch=True)
+        self.log('val_f1',f1,on_step=False,on_epoch=True)
+        self.log('val_recall',recall,on_step=False,on_epoch=True)
+                
+            
+    def configure_optimizers(self):
+        optimizer = torch.optim.Adam(self.model.parameters(), lr=self.learning_rate,betas=(self.beta1,self.beta2),eps=self.eps)
+        scheduler = lr_scheduler.StepLR(optimizer, step_size=10, gamma=0.1)
+        return {'optimizer': optimizer, 'lr_scheduler': scheduler}
+    
+    
+    #load model
+   
+    
+    
+    
+
+st.markdown("<h1 style='text-align: center; '>Brain Tumor Diagnosis</h1>",unsafe_allow_html=True)
+
+
+
+
+# Display a file uploader widget for the user to upload an image
+
+uploaded_file = st.file_uploader("Choose an Brain MRI image file", type=["jpg", "jpeg", "png"])
+
+# Load the uploaded image, or display emojis if no file was uploaded
+with st.container():
+    if uploaded_file is not None:
+        
+        image = Image.open(uploaded_file)
+        st.image(image, caption='Diagnosis', use_column_width=True)
+        model = timm.create_model(model_name='efficientnet_b1', pretrained=True,num_classes=4)
+        data_cfg = timm.data.resolve_data_config(model.pretrained_cfg)
+        transform = timm.data.create_transform(**data_cfg)
+        model_transforms = torchvision.transforms.Compose([transform])
+        transformed_image = model_transforms(image)
+        brain_model = torch.load('models/brain_model.pth')
+        
+        brain_model.eval()
+        with torch.inference_mode():
+            with st.progress(100):
+                
+                #class_names = ['Glinomia','Meningomia','notumar','pituary']
+                prediction = torch.nn.functional.softmax(brain_model(transformed_image.unsqueeze(dim=0))[0], dim=0)
+                prediction_score, pred_label_idx = torch.topk(prediction, 1)
+                pred_label_idx.squeeze_()
+                predicted_label = idx_to_labels[str(pred_label_idx.item())]
+                st.write( f'Predicted Label: {predicted_label}')
+                if st.button('Know More'):
+                    generator = pipeline("text-generation",model=text_model,tokenizer=tokenizer)
+                    input_text = f"Patient has {predicted_label} and is advised to take the following medicines:"
+                    with st.spinner('Generating Text'):
+                        generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)
+                    st.markdown(generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)[0]['generated_text'])
+                
+                    
+            
+            
+            
+            
+
+    
+        
+        
+    
+        
+        
+    else:
+        st.success("Please upload an image file 🧠")
+        
+        
+    
+    
+    
+## Model Explainibilty Dashboard using Captum 
+
+
+
+
+    
+
+

Medical Diagnosis App/pages/Chest.py

@@ -0,0 +1,163 @@
+import streamlit as st
+from PIL import Image
+import torch.nn as nn
+import timm
+import torch
+import time
+import torchmetrics
+from torchmetrics import F1Score,Recall,Accuracy
+import torch.optim.lr_scheduler as lr_scheduler
+import torchvision.models as models
+import lightning.pytorch as pl
+import torchvision
+from lightning.pytorch.loggers import WandbLogger
+import captum
+import matplotlib.pyplot as plt
+import json 
+from transformers import pipeline, set_seed
+from transformers import BioGptTokenizer, BioGptForCausalLM
+text_model = BioGptForCausalLM.from_pretrained("microsoft/biogpt")
+tokenizer = BioGptTokenizer.from_pretrained("microsoft/biogpt")
+labels_path = '/Users/vikram/Python/Medical Diagnosis App/labels.json'
+
+
+with open(labels_path) as json_data:
+    idx_to_labels = json.load(json_data)
+
+
+
+class FineTuneModel(pl.LightningModule):
+    def __init__(self, model_name, num_classes, learning_rate, dropout_rate,beta1,beta2,eps):
+        super().__init__()
+        self.model_name = model_name
+        self.num_classes = num_classes
+        self.learning_rate = learning_rate
+        self.beta1 = beta1
+        self.beta2 = beta2
+        self.eps = eps
+        self.dropout_rate = dropout_rate
+        self.model = timm.create_model(self.model_name, pretrained=True,num_classes=self.num_classes)
+        self.loss_fn = nn.CrossEntropyLoss()
+        self.f1 = F1Score(task='multiclass', num_classes=self.num_classes)
+        self.recall = Recall(task='multiclass', num_classes=self.num_classes)
+        self.accuracy = Accuracy(task='multiclass', num_classes=self.num_classes)
+        
+        #for param in self.model.parameters():
+            #param.requires_grad = True
+        #self.model.classifier= nn.Sequential(nn.Dropout(p=self.dropout_rate),nn.Linear(self.model.classifier.in_features, self.num_classes))
+        #self.model.classifier.requires_grad = True
+            
+
+    def forward(self, x):
+        return self.model(x)
+
+    def training_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('train_loss', loss,on_step=False,on_epoch=True)
+        self.log('train_acc', acc,on_step=False,on_epoch = True)
+        self.log('train_f1',f1,on_step=False,on_epoch=True)
+        self.log('train_recall',recall,on_step=False,on_epoch=True)
+        return loss
+            
+    def validation_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('val_loss', loss,on_step=False,on_epoch=True)
+        self.log('val_acc', acc,on_step=False,on_epoch=True)
+        self.log('val_f1',f1,on_step=False,on_epoch=True)
+        self.log('val_recall',recall,on_step=False,on_epoch=True)
+                
+            
+    def configure_optimizers(self):
+        optimizer = torch.optim.Adam(self.model.parameters(), lr=self.learning_rate,betas=(self.beta1,self.beta2),eps=self.eps)
+        scheduler = lr_scheduler.StepLR(optimizer, step_size=10, gamma=0.1)
+        return {'optimizer': optimizer, 'lr_scheduler': scheduler}
+    
+    
+    #load model
+   
+    
+    
+    
+
+st.markdown("<h1 style='text-align: center; '>Chest Xray Diagnosis</h1>",unsafe_allow_html=True)
+
+
+
+
+# Display a file uploader widget for the user to upload an image
+uploaded_file = st.file_uploader("Choose an Chest XRay Image file", type=["jpg", "jpeg", "png"])
+
+# Load the uploaded image, or display emojis if no file was uploaded
+if uploaded_file is not None:
+    
+    image = Image.open(uploaded_file)
+    st.image(image, caption='Diagnosis',width=224, use_column_width=True)
+    model = timm.create_model(model_name='efficientnet_b2', pretrained=True,num_classes=4)
+    data_cfg = timm.data.resolve_data_config(model.pretrained_cfg)
+    transform = timm.data.create_transform(**data_cfg)
+    model_transforms = torchvision.transforms.Compose([transform])
+    transformed_image = model_transforms(image)
+    xray_model = torch.load('models/timm_xray_model.pth')
+    
+    xray_model.eval()
+    
+    
+    
+    with torch.inference_mode():
+        with st.progress(100):
+            
+            prediction = torch.nn.functional.softmax(xray_model(transformed_image.unsqueeze(dim=0))[0], dim=0)
+            prediction_score, pred_label_idx = torch.topk(prediction, 1)
+            pred_label_idx.squeeze_()
+            predicted_label = idx_to_labels[str(pred_label_idx.item())]
+            st.write( f'Predicted Label: {predicted_label}')
+            if st.button('Know More'):
+                generator = pipeline("text-generation",model=text_model,tokenizer=tokenizer)
+                input_text = f"Patient has {predicted_label} and is advised to take the following medicines:"
+                with st.spinner('Generating Text'):
+                    generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)
+                st.markdown(generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)[0]['generated_text'])
+            
+            
+            
+            
+
+    
+        
+        
+    
+    
+    
+else:
+    st.success("Please upload an image file ⚕️")
+    
+    
+    
+    
+    
+
+
+
+
+
+
+
+
+
+
+
+
+
+    
+
+

Medical Diagnosis App/pages/Model Dashboard.py

@@ -0,0 +1,7 @@
+import streamlit as st
+
+st.components.v1.iframe(src = 'https://api.wandb.ai/links/vikramxd/nw5ru81j',width = 1000, height = 800,scrolling = True)
+
+
+
+

Medical Diagnosis App/pages/Skin.py

@@ -0,0 +1,145 @@
+import streamlit as st
+from PIL import Image
+import torch.nn as nn
+import timm
+import torch
+import torchmetrics
+from torchmetrics import F1Score,Recall,Accuracy
+import torch.optim.lr_scheduler as lr_scheduler
+import torchvision.models as models
+import lightning.pytorch as pl
+import torchvision
+from lightning.pytorch.loggers import WandbLogger
+import shap
+import matplotlib.pyplot as plt
+import json 
+from transformers import pipeline, set_seed
+from transformers import BioGptTokenizer, BioGptForCausalLM
+text_model = BioGptForCausalLM.from_pretrained("microsoft/biogpt")
+tokenizer = BioGptTokenizer.from_pretrained("microsoft/biogpt")
+labels_path = '/Users/vikram/Python/Medical Diagnosis App/skin_labels.json'
+from captum.attr import DeepLift , visualization
+
+with open(labels_path) as json_data:
+    idx_to_labels = json.load(json_data)
+
+
+
+class FineTuneModel(pl.LightningModule):
+    def __init__(self, model_name, num_classes, learning_rate, dropout_rate,beta1,beta2,eps):
+        super().__init__()
+        self.model_name = model_name
+        self.num_classes = num_classes
+        self.learning_rate = learning_rate
+        self.beta1 = beta1
+        self.beta2 = beta2
+        self.eps = eps
+        self.dropout_rate = dropout_rate
+        self.model = timm.create_model(self.model_name, pretrained=True,num_classes=self.num_classes)
+        self.loss_fn = nn.CrossEntropyLoss()
+        self.f1 = F1Score(task='multiclass', num_classes=self.num_classes)
+        self.recall = Recall(task='multiclass', num_classes=self.num_classes)
+        self.accuracy = Accuracy(task='multiclass', num_classes=self.num_classes)
+        
+        #for param in self.model.parameters():
+            #param.requires_grad = True
+        #self.model.classifier= nn.Sequential(nn.Dropout(p=self.dropout_rate),nn.Linear(self.model.classifier.in_features, self.num_classes))
+        #self.model.classifier.requires_grad = True
+            
+
+    def forward(self, x):
+        return self.model(x)
+
+    def training_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('train_loss', loss,on_step=False,on_epoch=True)
+        self.log('train_acc', acc,on_step=False,on_epoch = True)
+        self.log('train_f1',f1,on_step=False,on_epoch=True)
+        self.log('train_recall',recall,on_step=False,on_epoch=True)
+        return loss
+            
+    def validation_step(self, batch, batch_idx):
+        x, y = batch
+        y_hat = self.model(x)
+        loss = self.loss_fn(y_hat, y)
+        acc =  self.accuracy(y_hat.argmax(dim=1),y)
+        f1 = self.f1(y_hat.argmax(dim=1),y)
+        recall = self.recall(y_hat.argmax(dim=1),y)
+        self.log('val_loss', loss,on_step=False,on_epoch=True)
+        self.log('val_acc', acc,on_step=False,on_epoch=True)
+        self.log('val_f1',f1,on_step=False,on_epoch=True)
+        self.log('val_recall',recall,on_step=False,on_epoch=True)
+                
+            
+    def configure_optimizers(self):
+        optimizer = torch.optim.Adam(self.model.parameters(), lr=self.learning_rate,betas=(self.beta1,self.beta2),eps=self.eps)
+        scheduler = lr_scheduler.StepLR(optimizer, step_size=10, gamma=0.1)
+        return {'optimizer': optimizer, 'lr_scheduler': scheduler}
+    
+    
+    #load model
+   
+    
+    
+    
+
+st.markdown("<h1 style='text-align: center; '>Skin Leision Diagnosis</h1>",unsafe_allow_html=True)
+
+
+
+
+# Display a file uploader widget for the user to upload an image
+
+uploaded_file = st.file_uploader("Choose an Skin image file", type=["jpg", "jpeg", "png"])
+
+# Load the uploaded image, or display emojis if no file was uploaded
+with st.container():
+    if uploaded_file is not None:
+        
+        image = Image.open(uploaded_file)
+        st.image(image, caption='Diagnosis', use_column_width=True)
+        model = timm.create_model(model_name='efficientnet_b0', pretrained=True,num_classes=4)
+        data_cfg = timm.data.resolve_data_config(model.pretrained_cfg)
+        transform = timm.data.create_transform(**data_cfg)
+        model_transforms = torchvision.transforms.Compose([transform])
+        transformed_image = model_transforms(image)
+        brain_model = torch.load('models/timm_skin_model.pth')
+        
+        brain_model.eval()
+        with torch.inference_mode():
+            with st.progress(100):
+                
+                #class_names = ['Glinomia','Meningomia','notumar','pituary']
+                prediction = torch.nn.functional.softmax(brain_model(transformed_image.unsqueeze(dim=0))[0], dim=0)
+                prediction_score, pred_label_idx = torch.topk(prediction, 1)
+                pred_label_idx.squeeze_()
+                predicted_label = idx_to_labels[str(pred_label_idx.item())]
+                st.write( f'Predicted Label: {predicted_label}')
+                if st.button('Know More'):
+                    generator = pipeline("text-generation",model=text_model,tokenizer=tokenizer)
+                    input_text = f"Patient has {predicted_label} and is advised to take the following medicines:"
+                    with st.spinner('Generating Text'):
+                        generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)
+                    st.markdown(generator(input_text, max_length=300, do_sample=True, top_k=50, top_p=0.95, num_return_sequences=1)[0]['generated_text'])
+                
+                    
+            
+            
+            
+            
+
+    
+        
+        
+    
+        
+        
+    else:
+        st.success("Please upload an image file 🧠")
+        
+        

Medical Diagnosis App/requirements.txt

@@ -0,0 +1,11 @@
+captum==0.6.0
+lightning==2.0.1
+matplotlib==3.6.3
+Pillow==9.5.0
+shap==0.41.0
+streamlit==1.20.0
+timm==0.6.13
+torch==2.0.0
+torchmetrics==0.11.4
+torchvision==0.15.1
+transformers==4.27.4

Medical Diagnosis App/skin_labels.json

@@ -0,0 +1,10 @@
+{
+    "0":"actinic keratoses and intraepithelial carcinoma",
+    "1": "basal cell carcinoma",
+    "2":"benign keratosis-like lesions",
+    "3": "dermatofibroma",
+    "4": "melanoma",
+    "5": "melanocytic nevi", 
+    "6": "vascular lesions"
+
+}