I+NIR-SC-UFES: a multimodal image and Near-Infrared (NIR) spectroscopy dataset for skin cancer

Dataset Summary The collection of the Image & NIR Spectroscopy for Skin Cancer (I+NIR-SC-UFES) dataset took about one year and resulted in a multimodal dataset containing 774 samples with consistent image–spectrum pairs. The images were acquired using a smartphone under natural light, and the spectra were collected using near-infrared (NIR) spectroscopy. Data acquisition was conducted in partnership with the Dermatological and Surgical Assistance Program (PAD, in Portuguese: Programa de Assistência Dermatológica e Cirúrgica), a nonprofit program at the Federal University of Espírito Santo (UFES, Brazil). The PAD program focuses on providing free treatment for skin lesions, particularly benefiting low-income individuals in rural areas. The dataset encompasses images of six distinct skin lesion categories: - Basal Cell Carcinoma (BCC), Squamous Cell Carcinoma (SCC), and Melanoma (MEL), corresponding to biopsy-confirmed skin cancers; - Actinic Keratosis (ACK), Seborrheic Keratosis (SEK), and Nevus (NEV), corresponding to non-cancerous skin lesions. The .csv file contains 125 spectral values for each sample, with each value representing the average measurement over a 6.4 nm interval across the wavelength range from 900 to 1700 nm. The file also includes the identifier of the corresponding image, which is stored in the imgs directory. All images are provided in .jpg format. Ethics statement The dataset was collected along with the Dermatological and Surgical Assistance Program (PAD) of the Federal University of Espírito Santo. The program is managed by the Department of Specialized Medicine and was approved by the university ethics committee (nº 500002/478) and the Brazilian government through Plataforma Brasil (nº 4.007.097), the Brazilian agency responsible for research involving human beings. In addition, all data is collected under patient consent and the patient’s privacy is completely preserved.数据集概述 本次采集的皮肤癌图像与近红外光谱数据集（Image & NIR Spectroscopy for Skin Cancer，缩写I+NIR-SC-UFES）历时约一年，最终构建得到包含774组匹配一致的图像-光谱对的多模态数据集。 图像采用智能手机在自然光环境下采集，光谱数据则通过近红外（near-infrared, NIR）光谱仪获取。本次数据采集工作与巴西圣埃斯皮里图联邦大学（Federal University of Espírito Santo, UFES）下属的皮肤科与外科援助项目（葡萄牙语全称：Programa de Assistência Dermatológica e Cirúrgica，缩写PAD）合作完成。该项目为非营利性医疗项目，专注于为皮肤病变患者提供免费诊疗服务，尤其惠及农村地区的低收入人群。 本数据集涵盖六类不同的皮肤病变图像： - 基底细胞癌（Basal Cell Carcinoma, BCC）、鳞状细胞癌（Squamous Cell Carcinoma, SCC）与黑色素瘤（Melanoma, MEL），均为经活检证实的皮肤恶性肿瘤； - 光化性角化病（Actinic Keratosis, ACK）、脂溢性角化病（Seborrheic Keratosis, SEK）与痣（Nevus, NEV），均为良性皮肤病变。 配套的.csv文件包含每组样本的125条光谱数值，每条数值对应900 nm至1700 nm波长范围内、6.4 nm区间的平均测量结果。该文件同时收录了对应图像的标识符，相关图像存储于imgs目录下，所有图像均采用.jpg格式。 伦理声明 本数据集的采集工作依托巴西圣埃斯皮里图联邦大学皮肤科与外科援助项目（PAD）开展。该项目由该校专科医学部管理，已通过该校伦理委员会审批（编号：500002/478），并经巴西政府通过负责人类受试者研究的官方机构Plataforma Brasil审核通过（编号：4.007.097）。此外，所有数据均在获得患者知情同意后采集，且严格保障患者隐私。