Dataset Viewer
Auto-converted to Parquet Duplicate
Image Index
stringlengths
16
16
Texts
stringlengths
84
210
View Position
stringclasses
2 values
Image Features
sequencelengths
512
512
Text Features
sequencelengths
512
512
Atelectasis
int32
0
1
Cardiomegaly
int32
0
1
Effusion
int32
0
1
Infiltration
int32
0
1
Mass
int32
0
1
Nodule
int32
0
1
Pneumonia
int32
0
1
Pneumothorax
int32
0
1
Consolidation
int32
0
1
Edema
int32
0
1
Emphysema
int32
0
1
Fibrosis
int32
0
1
Hernia
int32
0
1
Pleural_Thickening
int32
0
1
No_Finding
int32
0
1
00000001_000.png
This photo of a chest x-ray shows a Cardiomegaly finding. The image is taken from a PA view.
PA
[ -0.047554731369018555, -0.024417875334620476, -0.2809467911720276, -0.025858107954263687, -0.014165667816996574, 0.003804249921813607, -0.023984646424651146, 0.016693774610757828, -0.0011217523133382201, -0.021765759214758873, 0.04472145810723305, 0.01795242913067341, -0.029904279857873917, ...
[ -0.014336158521473408, -0.051427945494651794, -0.03088880330324173, -0.025827940553426743, -0.019897861406207085, 0.016288522630929947, 0.011414722539484501, -0.0012368570314720273, 0.02699963003396988, -0.05022412911057472, -0.011788640171289444, 0.011947030201554298, -0.018385400995612144,...
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
00000001_001.png
This photo of a chest x-ray shows multiple findings including Cardiomegaly and Emphysema. The image is taken from a PA view.
PA
[ -0.04611440375447273, -0.0379432812333107, -0.28582221269607544, -0.019533470273017883, -0.023110153153538704, 0.01052901428192854, -0.033620916306972504, 0.008291111327707767, -0.007207818329334259, -0.026053784415125847, 0.029763756319880486, 0.02085862122476101, -0.03322592377662659, 0....
[ -0.03448459878563881, -0.05419719219207764, -0.03807877376675606, -0.030541833490133286, -0.011128371581435204, 0.020796822383999825, -0.0021311778109520674, -0.01328555028885603, 0.022600580006837845, -0.01882605440914631, 0.008783583529293537, -0.0052074259147048, -0.013202871195971966, ...
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
00000001_002.png
This photo of a chest x-ray shows multiple findings including Cardiomegaly and Effusion. The image is taken from a PA view.
PA
[ -0.046720653772354126, -0.040042001754045486, -0.29038652777671814, -0.028128746896982193, -0.029858386144042015, 0.03533126786351204, -0.033193349838256836, 0.009468233212828636, 0.0028854617848992348, -0.024626603350043297, 0.024211443960666656, 0.02468465454876423, -0.012263563461601734, ...
[ -0.02680680900812149, -0.05335159972310066, -0.041031938046216965, -0.008345304988324642, -0.019419901072978973, 0.057277169078588486, -0.02209486998617649, -0.030611135065555573, 0.031314410269260406, -0.04379505664110184, -0.028816401958465576, 0.01494523137807846, -0.02790152095258236, ...
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
00000002_000.png
This photo of a chest x-ray shows no findings available. The image is taken from a PA view.
PA
[ -0.05449114367365837, -0.002965022111311555, -0.2812284827232361, -0.020111070945858955, -0.016491612419486046, 0.0380941778421402, -0.012179896235466003, 0.03852112218737602, -0.01537646446377039, 0.004737014416605234, 0.01630519889295101, -0.007682950235903263, -0.0011849789880216122, -0...
[ -0.0198361799120903, -0.040187571197748184, -0.03979671001434326, 0.006723976694047451, -0.0030163845513015985, 0.009550536051392555, 0.005970212165266275, 0.011782718822360039, 0.0056159100495278835, -0.02836361713707447, -0.003135020611807704, -0.019343528896570206, -0.02240701951086521, ...
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
00000003_001.png
This photo of a chest x-ray shows a Hernia finding. The image is taken from a PA view.
PA
[ -0.0029371855780482292, 0.03051556833088398, -0.2535656690597534, -0.025174815207719803, -0.0022402536123991013, 0.026561236009001732, 0.0007812470430508256, -0.00048237049486488104, -0.021610625088214874, -0.0076167527586221695, 0.020935960114002228, -0.022095663473010063, -0.03685499727725...
[ -0.02361217513680458, 0.0213457178324461, -0.031623583287000656, -0.01281907968223095, -0.023813370615243912, 0.055228397250175476, 0.0399709977209568, 0.02986587956547737, -0.00025465746875852346, -0.04409777745604515, 0.01704263500869274, 0.023485258221626282, -0.026028048247098923, -0.0...
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
00000003_002.png
This photo of a chest x-ray shows a Hernia finding. The image is taken from a PA view.
PA
[ -0.022250283509492874, 0.03808863088488579, -0.2374359369277954, -0.01571289636194706, 0.020297953858971596, 0.04692203924059868, 0.015179627574980259, 0.022372310981154442, -0.04126177728176117, 0.0019121458753943443, 0.02253575064241886, -0.010380223393440247, -0.05097575858235359, -0.07...
[ -0.02361217513680458, 0.0213457178324461, -0.031623583287000656, -0.01281907968223095, -0.023813370615243912, 0.055228397250175476, 0.0399709977209568, 0.02986587956547737, -0.00025465746875852346, -0.04409777745604515, 0.01704263500869274, 0.023485258221626282, -0.026028048247098923, -0.0...
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
00000003_003.png
This photo of a chest x-ray shows multiple findings including Hernia and Infiltration. The image is taken from a PA view.
PA
[ -0.021311284974217415, 0.0355827771127224, -0.2669873535633087, -0.013546047732234001, -0.0018805214203894138, 0.004954432602971792, 0.012503527104854584, 0.010205771774053574, -0.024577399715781212, -0.008344548754394054, 0.01751410961151123, -0.0035172258503735065, -0.04443413391709328, ...
[ -0.03479829803109169, 0.030155794695019722, -0.0435178168118, -0.0272881630808115, -0.008127078413963318, 0.07979650050401688, 0.03583281859755516, 0.013478685170412064, -0.014633982442319393, -0.02047635242342949, 0.029456809163093567, 0.026505662128329277, -0.026806427165865898, -0.05781...
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
00000003_004.png
This photo of a chest x-ray shows a Hernia finding. The image is taken from a PA view.
PA
[ -0.02961430326104164, 0.013943638652563095, -0.28719431161880493, -0.016743140295147896, 0.007997116073966026, 0.031386133283376694, -0.009863189421594143, 0.035565316677093506, -0.026128582656383514, -0.006340420804917812, 0.0010202032281085849, -0.01986430399119854, -0.031731944531202316, ...
[ -0.02361217513680458, 0.0213457178324461, -0.031623583287000656, -0.01281907968223095, -0.023813370615243912, 0.055228397250175476, 0.0399709977209568, 0.02986587956547737, -0.00025465746875852346, -0.04409777745604515, 0.01704263500869274, 0.023485258221626282, -0.026028048247098923, -0.0...
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
00000003_005.png
This photo of a chest x-ray shows a Hernia finding. The image is taken from a PA view.
PA
[-0.03054821863770485,0.031076371669769287,-0.2557828426361084,-0.0308160949498415,-0.00539532024413(...TRUNCATED)
[-0.02361217513680458,0.0213457178324461,-0.031623583287000656,-0.01281907968223095,-0.0238133706152(...TRUNCATED)
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
00000003_006.png
This photo of a chest x-ray shows a Hernia finding. The image is taken from a PA view.
PA
[-0.0376569963991642,0.020924098789691925,-0.25126561522483826,-0.035456590354442596,0.0032629261258(...TRUNCATED)
[-0.02361217513680458,0.0213457178324461,-0.031623583287000656,-0.01281907968223095,-0.0238133706152(...TRUNCATED)
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
End of preview. Expand in Data Studio

NIH-CXR14-BiomedCLIP-Features Dataset

This dataset is derived from the NIH Chest X-ray Dataset (NIH-CXR14) and processed using the BiomedCLIP-PubMedBERT_256-vit_base_patch16_224 model from Microsoft. It contains image and text features extracted from chest X-ray images and their corresponding textual findings.

Dataset Description

The original NIH-CXR14 dataset comprises 112,120 chest X-ray images with disease labels from 30,805 unique patients. This processed dataset includes:

  • Image Features: Extracted using the vision encoder of BiomedCLIP (512 dimensions).
  • Text Features: Extracted using the text encoder of BiomedCLIP (512 dimensions).
  • Finding Labels: The original disease labels, processed and converted into a multi-label format.
  • Image Index: Unique identifiers for each image.
  • View Position: The view position of the X-ray (e.g., PA, AP).
  • Processed Text: A grammatically correct text prompt generated from the finding labels, designed for use with the BiomedCLIP model.

Processing Steps

  1. Data Loading: The original NIH-CXR14 image and text datasets were loaded.
  2. Text Preprocessing:
    • Problematic characters (|) were replaced with commas.
    • "No Finding" labels were converted to "No_Finding".
    • Finding labels were split into individual findings.
    • Grammatically correct text prompts were generated based on the finding labels and view position.
  3. Feature Extraction:
    • Images and text prompts were preprocessed using the BiomedCLIP preprocessors.
    • Image and text features were extracted using the BiomedCLIP model.
  4. Data Storage:
    • Extracted features, image indices, view positions, raw texts, and finding labels were stored in Parquet files.
    • The dataset was chunked into multiple Parquet files for efficient storage and retrieval.

Dataset Structure

The dataset is organized into Parquet files, each containing the following columns:

  • Image Index: String, unique identifier for each image.
  • Image Features: List of floats, image features extracted by BiomedCLIP.
  • Text Features: List of floats, text features extracted by BiomedCLIP.
  • View Position: String, view position of the X-ray.
  • Texts: String, processed text prompts.
  • [Finding Label]: Integer (0 or 1), multi-label representation of each finding.

Usage

This dataset can be used for various tasks, including:

  • Multi-label classification: Using the extracted features to predict disease findings.
  • Retrieval: Retrieving relevant X-ray images based on text queries or vice versa.
  • Fine-tuning: Fine-tuning models for medical image analysis tasks.

Installation

To load the dataset, you can use the datasets library from Hugging Face:

from datasets import load_dataset

dataset = load_dataset("Yasintuncer/NIH-CXR14-BiomedCLIP-Features")
Downloads last month
38