ERCPMP: an endoscopic image and video dataset for colorectal polyps morphology and pathology

Abstract

This dataset contains demographic, morphological and pathological data, endoscopic images and videos of 191 patients with colorectal polyps. Morphological data is included based on the latest international gastroenterology classification references such as Paris, Pit and JNET classification. Pathological data includes the diagnosis of the polyps including Tubular, Villous, Tubulovillous, Hyperplastic, Serrated, Inflammatory and Adenocarcinoma with Dysplasia Grade & Differentiation.

Objectives: Today the most important challenge of developing accurate algorithms for medical prediction, detection, diagnosis, treatment and prognosis is data. ERCPMP is an Endoscopic Image and Video Dataset for Recognition of Colorectal Polyps Morphology and Pathology. This dataset can be used for developing deep learning algorithms for polyps detection, classification, and segmentation.

Data description: Images were captured with Olympus colonoscope and are presented in RGB format, JPG type with the resolution of 368 * 256 pixels and 96 dpi. The name of each file (image or video) includes pathological diagnosis, grade and JNet classification of the related polyp.

Objective

Colorectal cancer (CRC) is a significant cause of mortality worldwide, responsible for an estimated 1.9 million new cases and 935,000 deaths globally among 5.2 million diagnosed cases in 2020 [1]. It is the third most prevalent malignancy worldwide and the second major cause of cancer-related mortality [1]. Detecting CRC early through screening methods like colonoscopy, fecal occult blood tests, and sigmoidoscopy is crucial for improving patient outcomes, which can detect polyps and early-stage malignancies that can be excised before they progress [2, 3].

Colorectal polyps are atypical growths found in the colon or rectum, often discovered during routine colonoscopy exams [4]. Most CRCs develop from precancerous adenomatous polyps [4, 5]. It has been demonstrated that early diagnosis and excision of precancerous colorectal polyps dramatically reduces the risk of colorectal cancer. The excision of such polyps during colonoscopy can prevent the development of cancer from these lesions [3].

In recent times, considerable endeavors have been undertaken to anticipate and identify various forms of cancer by utilizing artificial intelligence (AI) and its subfields, such as machine learning and deep learning [6,7,8]. The initial crucial phase towards accomplishing this objective involves obtaining an appropriate dataset. Consequently, this study sought to create a meticulously structured collection of images and videos encompassing demographic information, histopathological attributes (including grading, differentiation, and diagnosis), and morphological characteristics (such as size, circumference, Paris class, Pit pattern, JNET classification, and LST type) of colorectal polyps.

ERCPMP [9] is a histopathological and morphological image and video dataset of 191 patients diagnosed with colorectal polyps including 796 images and 21 videos in total. These numbers are related to the current version, but it is under development to bring more data in the next versions. For queries regarding the latest updates and more information about this dataset, please refer to: https://databiox.com

Data description

ERCPMP [9] is the name of the prepared image and video dataset of this research. This is a morphological and histopathological image and video dataset of 191 patients diagnosed with colorectal polyps including 796 images and 21 videos in total. These numbers are related to the current version, but it is under development to bring more data in the next versions. Images were captured with Olympus colonoscope and are presented in RGB format, JPG type with the resolution of 368 * 256 pixels and 96 dpi. Videos were captured with the same device and are presented in MP4 type. An overview of data files is presented in Table 1, and a summary of technical information of the dataset is introduced in the supplementary file. File names in the dataset are based on patient codes provided in the accompanied excel file. The excel file contains anonymized information on each patient’s demographic data in addition to each polyp morphological and histopathological labeling.

Images were captured with Olympus colonoscope and are presented in RGB format, JPG type with the resolution of 368 * 256 pixels and 96 dpi. The name of each file (image or video) includes pathological diagnosis, grade and JNet classification of the related polyp.

The ERCPMP dataset [9] is distinct from similar datasets due to the presence of both morphological and histopathological characteristics of colorectal polyps in addition to including more polyp samples with various features. Generally, six necessary steps were implemented to arrange this dataset as listed:

• 1. Patients with colorectal polyps were diagnosed and their demographics were recorded.

• 2. Polyp anatomical features, morphology features, and surface pattern were assessed and classified.

• 3. Polyp samples were referred for histopathologic assessment.

• 4. Histologic diagnosis and grading were recorded.

• 5. A written informed consent was obtained from patients to include their clinical details.

• 6. The dataset was organized.

Table 1 Overview of data files/data sets

Limitations

The limitations of the study were the lack of access to reports, photos, and especially videos, and pathology reports of some patients that had been done in the past. Another limitation was the unpredictability of the pathological type of polyps, which caused asymmetry in the number of different pathological types of polyps.