baichuan-inc/OpenMM_Medical

OpenMM-Medical

Introduction

OpenMM-Medical is a comprehensive medical evaluation dataset, which is an integration of existing datasets. OpenMM-Medical spans multiple domains, including Magnetic Resonance Imaging (MRI), CT scans, X-rays, microscopy images, endoscopy, fundus imaging, and dermoscopy.

Components	Content	Type	Number	Metrics
ACRIMA	Fundus Photography	Multiple Choice Question Answering	159	Acc
Adam Challenge	Endoscopy	Multiple Choice Question Answering	87	Acc
ALL Challenge	Microscopy Images	Multiple Choice Question Answering	342	Acc
BioMediTech	Microscopy Images	Multiple Choice Question Answering	511	Acc
Blood Cell	Microscopy Images	Multiple Choice Question Answering	1175	Acc
BreakHis	Magnetic Resonance Imaging	Multiple Choice Question Answering	735	Acc
Chest CT Scan	CT Imaging	Multiple Choice Question Answering	871	Acc
Chest X-Ray PA	X-Ray	Multiple Choice Question Answering	850	Acc
CoronaHack	X-Ray	Multiple Choice Question Answering	684	Acc
Covid CT	CT Imaging	Multiple Choice Question Answering	199	Acc
Covid-19 tianchi	X-Ray	Multiple Choice Question Answering	96	Acc
Covid19 heywhale	X-Ray	Multiple Choice Question Answering	690	Acc
COVIDx CXR-4	X-Ray	Multiple Choice Question Answering	485	Acc
CRC100k	Magnetic Resonance Imaging	Multiple Choice Question Answering	1322	Acc
DeepDRiD	Fundus Photography	Multiple Choice Question Answering	131	Acc
Diabetic Retinopathy	Fundus Photography	Multiple Choice Question Answering	2051	Acc
DRIMDB	Fundus Photography	Multiple Choice Question Answering	132	Acc
Fitzpatrick 17k	Dermoscopy	Multiple Choice Question Answering	1552	Acc
HuSHeM	Microscopy Images	Multiple Choice Question Answering	89	Acc
ISBI2016	Dermoscopy	Multiple Choice Question Answering	681	Acc
ISIC2018	Dermoscopy	Multiple Choice Question Answering	272	Acc
ISIC2019	Dermoscopy	Multiple Choice Question Answering	1952	Acc
ISIC2020	Dermoscopy	Multiple Choice Question Answering	1580	Acc
JSIEC	Fundus Photography	Multiple Choice Question Answering	220	Acc
Knee Osteoarthritis	X-Ray	Multiple Choice Question Answering	518	Acc
MAlig Lymph	Magnetic Resonance Imaging	Multiple Choice Question Answering	149	Acc
MHSMA	Microscopy Images	Multiple Choice Question Answering	1282	Acc
MIAS	X-Ray	Multiple Choice Question Answering	142	Acc
Monkeypox Skin Image 2022	Dermoscopy	Multiple Choice Question Answering	163	Acc
Mura	X-Ray	Multiple Choice Question Answering	1464	Acc
NLM- Malaria Data	Magnetic Resonance Imaging	Multiple Choice Question Answering	75	Acc
OCT & X-Ray 2017	X-Ray, Optical Coherence Tomography	Multiple Choice Question Answering	1301	Acc
OLIVES	Fundus Photography	Multiple Choice Question Answering	593	Acc
PAD-UFES-20	Dermoscopy	Multiple Choice Question Answering	479	Acc
PALM2019	Fundus Photography	Multiple Choice Question Answering	510	Acc
Pulmonary Chest MC	X-Ray	Multiple Choice Question Answering	38	Acc
Pulmonary Chest Shenzhen	X-Ray	Multiple Choice Question Answering	296	Acc
RadImageNet	CT; Magnetic Resonance Imaging; Ultrasound	Multiple Choice Question Answering	56697	Acc
Retinal OCT-C8	Optical Coherence Tomography	Multiple Choice Question Answering	4016	Acc
RUS CHN	X-Ray	Multiple Choice Question Answering	1982	Acc
SARS-CoV-2 CT-scan	CT	Multiple Choice Question Answering	910	Acc
Yangxi	Fundus Photography	Multiple Choice Question Answering	1515	Acc

Usage

The following steps detail how to use Baichuan-Omni-1.5 with OpenMM-Medical for evaluation using VLMEvalKit:

---

1. Add baichuan.py in VLMEvalKit/vlmeval/vlm

Download baichuan.py (which defines the Baichuan model class) and add it in VLMEvalKit/vlmeval/vlm.

---

2. Modify VLMEvalKit/vlmeval/vlm/__init__.py

Add the following line:

from .baichuan import Baichuan

---

3. Modify VLMEvalKit/vlmeval/config.py

Import the Baichuan model:

from vlmeval.vlm import Baichuan

Add the Baichuan-omni model configuration:

'Baichuan-omni': partial(
    Baichuan, 
    sft=True, 
    model_path='/your/path/to/the/model/checkpoint'
)

---

4. Modify VLMEvalKit/vlmeval/dataset/image_mcq.py

Download image_mcq.py and add the following code to define the OpenMMMedical class. Ensure the image_folder points to your OpenMM-Medical dataset location:

class OpenMMMedical(ImageMCQDataset):

    @classmethod
    def supported_datasets(cls):
        return ['OpenMMMedical']

    def load_data(self, dataset='OpenMMMedical'):
        image_folder = "/your/path/to/OpenMM_Medical"
        def generate_tsv(pth):
            import csv
            from pathlib import Path
            tsv_file_path = os.path.join(LMUDataRoot(), f'{dataset}.tsv')
        ...

---

5. Update VLMEvalKit/vlmeval/dataset/__init__.py

Import OpenMMMedical:

from .image_mcq import (
    ImageMCQDataset, MMMUDataset, CustomMCQDataset, 
    MUIRDataset, GMAIMMBenchDataset, MMERealWorld, OpenMMMedical
)

IMAGE_DATASET = [
    ImageCaptionDataset, ImageYORNDataset, ImageMCQDataset, ImageVQADataset,
    MathVision, MMMUDataset, OCRBench, MathVista, LLaVABench, MMVet,
    MTVQADataset, TableVQABench, MMLongBench, VCRDataset, MMDUDataset,
    DUDE, SlideVQA, MUIRDataset, GMAIMMBenchDataset, MMERealWorld, OpenMMMedical
]

---

6. Update VLMEvalKit/vlmeval/dataset/image_base.py

Modify the img_root_map function:

def img_root_map(dataset):
    if 'OpenMMMedical' in dataset:
        return 'OpenMMMedical'
    if 'OCRVQA' in dataset:
        return 'OCRVQA'
    if 'COCO_VAL' == dataset:
        return 'COCO'
    if 'MMMU' in dataset:
        return 'MMMU'

---

7. Run the Evaluation

Execute the following command to start the evaluation:

python run.py --data OpenMMMedical --model Baichuan-omni --verbose

---

Notes:

• Ensure that all paths (e.g., /your/path/to/OpenMM_Medical) are correctly specified.
• Confirm that the Baichuan model checkpoint is accessible at the defined model_path.
• Validate the dependencies and configurations of VLMEvalKit to avoid runtime issues.

With this setup, you should be able to evaluate OpenMM-Medical using Baichuan-Omni successfully.