How Data Labeling Tools Empower Bone Age Assessment
Bone age is the main method for evaluating the biological age of adolescents and children. Bone age and actual age are not consistent. Bone age examination can understand the development of bones, determine whether bone age is stunted or too fast, and understand the causes and prognosis of short stature. It has a wide range of uses in the fields of clinical medicine, forensic medicine and sports medicine.
At present, the commonly used bone age evaluation method in China is the China 05 Bone Age Standard, including TW3-C RUS, TW3-C Carpal and RUS-CHN methods. No matter what kind of evaluation method is used, it has higher requirements on the evaluation physician’s reading experience and working years. In 2014, the framers of the China 05 Bone Age Standard performed a reliability test of bone age readings on several physicians. The results showed that working years, experience in reading images, and whether they had participated in technical training had a greater impact on the accuracy of physicians’ bone age judgment.
And, a notable feature of bone age evaluation is that it requires both geographical and spatial dimensions as well as temporal dimensions. E.g:
● The bone age standards of developed countries such as Europe and the United States are not applicable to East Asia (China, Japan, South Korea).
● In the past 20 years, with the great changes in China’s social economy, the growth and development of Chinese children has shown a trend of significant acceleration. This makes bone age research and evaluation standards also follow the development and changes of the times.
It shows that the bone age data is not universal in the world, and doctors need to participate in training frequently to adapt to new changes. Whether from the perspective of physicians or from the perspective of data changes, if localized bone age data can be collected and bone age research and judgment based on AI technology, the accuracy of the application model will be more accurate and “reliable” than manual reading.
Shujiajia Pro Labeling Platform — Bone Age Labeling Tools
Datatang’s Sujijia Pro labeling platform focuses on bone age labeling in the field of medical labeling, integrates dicom medical data reading capabilities with bone age labeling, and launches bone age labeling templates, which support various bone age labeling methods, such as TW3-C RUS, TW3-C Carpal and RUS-CHN, and pay more attention to the efficiency and experience in bone age labeling scenarios.
It shows that for the classification and labeling of medical data, especially for the labeling scenarios that require detailed description of the classification label, the bone age labeling tool is also suitable and naturally has a friendly display effect.
In addition to annotating bone age, the bone age labeling tool also supports object detection labeling of medical images. Project managers can flexibly configure labels and attributes required by business on the tool page. Labeling physicians do not need to pay attention to configuration details, but only need to make a rectangular frame and choose labels on the labeling page.
At the same time, the bone age labeling tool also supports key point labeling. Project manager can configure the key point group in advance through the template. The labeling doctor only needs to read the film and label the key points in sequence. The key points are automatically switched in order to improve the labeling efficiency.
In terms of medical data format, Shujiajia Pro platform supports dataset management and annotation in DICOM medical digital imaging format and common 2D visible light image format. It also supports multiple data connection methods such as API push and private cloud import, achieving online reading and data viewing, and supports custom hiding of sensitive information such as patient names to meet sensitive information processing requirements and relevant privacy laws and regulations.
While developing the Shujiajia Pro annotation platform for medical scenarios, Datatang pays special attention to and abides by relevant national, local and industry laws and regulations.
In the future, the Shujiajia Pro labeling platform will further conduct comprehensive inspections and guide the development of platform functions in accordance with the laws and regulations. And based on the needs of customers, the following functions are planned to be launched to ensure the legal compliance of the platform, while providing more abundant functions to meet the medical data labeling needs.
● Data fitting and arbitration mechanism: Support multi-standard fitting arbitration process, and support cross-arbitration and third-party arbitration for objection data to ensure data quality.
● Repeatability accuracy index labeling evaluation: Support the method of subbed verification to count whether the results of multiple annotations of the same data by the same annotator in different time periods are consistent, and calculate the proportion of duplicate labels in the duplicate label samples.
● Accuracy index labeling evaluation: multi-dimensional evaluation is carried out by means of quality control personnel’s return rate, random sampling and statistical accuracy.
● Online training function: support junior annotators to complete training through online exercises and exams, and the system supports automatic scoring of submitted exam annotation data.
Founded in 2011, Datatang is a professional AI data service provider and committed to providing high-quality training data and data services for global AI companies. Relying on own data resources, technical advantages and intensive data processing experiences, Datatang provides data services to 1,000+ companies and institutions worldwide.
If you need data services, please feel free to contact us: email@example.com