Validation of AI Lumbar Spine Anatomy Measureing Function

RadiRad Co., Ltd.150 enrolled

Overview

Spinal degeneration and its associated clinical diseases are common ailments in aging societies. With the advent of a super-aging society, the importance of assistive technologies for spinal image interpretation is increasingly significant to enhance care efficiency and reduce medical personnel expenditure. Recently, due to the rapid development of artificial intelligence (AI) algorithm, AI-based computer-assisted detection (CADe) devices gradiually become a convenient method for spinal anatomy measurement. However, the accuracy of these devices has not been fully established. This study aims to validate the performance of RadiSpine (an application program) in spinal anatomy segmentation and measurement.

Study Type

OBSERVATIONAL

Enrollment

150

Conditions

Spinal Cord Compression

Eligibility

Sex: ALLMin age: 22 YearsMax age: 75 YearsHealthy volunteers:

Medical Language ↔ Plain English

Inclusion Criteria: The subjects should be aged 20 or older and younger than 75, with an equal gender distribution of 50% male and 50% female. From this group, 150 subjects with reasonable datavalues will be selected, with a requirement that at least 30% of them are male and at least 30% are female. Exclusion Criteria: 1. With history of spinal surgery 2. Spinal trauma 3. Spinal osteoporosis 4. Spinal metastasis or infection

Locations (1)

Taipei Veterans General Hospital

Outcomes

Primary Outcomes

Segmentation accuracy (Mean)

The minimum Mean Dice Coefficient (MDC), defined as the lower limit of the 95% confidence interval (CI) for MDC, is above a predetermined allowable limit equal to 0.8

Time frame: 30 mins per individual

Secondary Outcomes

Measurement accuracy

The maximum Mean Absolute Error (MAE), defined as the upper limit of the 95% CI for MAE, is below a predetermined allowable error limit equal to 2 mm

Time frame: 30 mins per individual