{"id":7702,"date":"2026-09-11T10:09:01","date_gmt":"2026-09-11T08:09:01","guid":{"rendered":"https:\/\/www.scs.ch\/uncategorized-en-us\/ai-detects-hip-malalignments\/"},"modified":"2026-09-20T07:38:58","modified_gmt":"2026-09-20T05:38:58","slug":"ai-detects-hip-malalignments","status":"publish","type":"post","link":"https:\/\/www.scs.ch\/en\/case-studies\/ai-detects-hip-malalignments\/","title":{"rendered":"AI Detects Hip Malalignments"},"content":{"rendered":"

Background<\/h2><\/div>

Approximately 3 percent of all newborns are born with hip malalignment. If detected early via ultrasound, it can usually be treated easily, for example with an abduction brace for a few weeks. If it remains undetected, this hip dysplasia leads to hip surgery before the age of 60 in 30 percent of those affected. The Graf alpha angle is particularly decisive for the assessment. However, its determination depends on the placement of the measurement lines and can vary between evaluators. The existing clinical data also showed a noticeable clustering of measured values exactly at the threshold of 60 degrees.<\/p>\n<\/div>

\"S\u00e4ugling<\/span><\/div><\/div><\/div>

Project Overview<\/h3><\/div>

Field of application<\/strong>
\nPediatrics, ultrasound diagnostics of the infant hip<\/p>\n<\/div>

<\/div><\/div>

Area<\/strong>
Medical & Life Science<\/p>\n<\/div>

<\/div><\/div>

Solution<\/strong>
AI-supported assessment of infant hips in ultrasound.<\/p>\n<\/div>

<\/div><\/div>

System environment <\/strong>
\nUltrasound images of the infant hip, Graf alpha and beta angles<\/p>\n<\/div>

<\/div><\/div>

Partners<\/strong>
\nDr. Stefan Essig, Dr. Thomas Baumann (Pediatricians)<\/p>\n<\/div>

<\/div><\/div>

Evidence<\/strong>
\nJoint study, scientifically published in Ultraschall in der Medizin<\/p>\n<\/div><\/div><\/div><\/div><\/div>

The SCS Contribution<\/h2><\/div>

SCS trained various algorithms using ultrasound images to determine the hip angle objectively and reproducibly, and proved the approach in a joint study with specialists and pediatricians.<\/p>\n<\/div><\/div><\/div>

<\/i><\/div>

Angle Objectively Determined<\/h4><\/div>
<\/div>
\n

A deep neural network predicts the lines in the ultrasound image from which the hip angle is derived, analogous to the medical procedure.<\/p>\n<\/div><\/div><\/div>

<\/i><\/div>

Database Cleaned<\/h4><\/div>
<\/div>
\n

For the training, the database was first cleaned by removing the excess of borderline cases.<\/p>\n<\/div><\/div><\/div>

<\/i><\/div>

Scientifically Verified<\/h4><\/div>
<\/div>
\n

SCS conducted a study together with specialists and pediatricians and compared the result with data from pediatricians that were not used for the training.<\/p>\n<\/div><\/div><\/div>

<\/i><\/div>

Image Quality Additionally Supported<\/h4><\/div>
<\/div>
\n

In addition to automated angle determination, the solution can alert physicians early to unsuitable ultrasound recordings and thus support the examination workflow.<\/p>\n<\/div><\/div><\/div>

<\/div><\/div><\/div><\/div>
\"Balkendiagramm<\/span><\/div><\/div><\/div><\/div><\/div>
<\/div>

Result<\/h2><\/div>

The comparison with independent data from pediatricians shows that the algorithm determines the hip angle reliably. The joint study confirmed that the angle can be determined objectively and reproducibly using deep learning. More precise and reproducible angle determination can support early diagnosis and alert physicians to unsuitable recordings at an early stage.<\/p>\n<\/div><\/div><\/div>

approx. 3 percent<\/h3><\/div>

of newborns are born with hip malalignment.<\/p>\n<\/div><\/div><\/div>

30 percent<\/h3><\/div>

of undetected cases lead to hip surgery before the age of 60.<\/p>\n<\/div><\/div><\/div>

scientifically published<\/h3><\/div>

Study results in Ultraschall in der Medizin.<\/p>\n<\/div><\/div><\/div><\/div><\/div>

Services Utilized<\/h2><\/div><\/div><\/div>

Algorithms, ML & AI<\/h3><\/div><\/div><\/div><\/div><\/div><\/div>

Computer Vision<\/h3><\/div><\/div><\/div><\/div><\/div><\/div>

Medical & Life Science<\/h3><\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div>
\"Smiling<\/div><\/div>

Your Contact Person<\/h2><\/div>
Beat H\u00f6rmann<\/div><\/div>

Healthtech & Innovation
\n+41 43 456 16 00<\/a><\/p>\n<\/div>

Speak with the responsible specialist<\/span>Speak with the responsible specialist<\/span><\/div><\/a><\/div>

For questions regarding machine learning, medical image analysis, and objective, reproducible diagnostic support.<\/p>\n<\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div>

Further Case Studies<\/h2><\/div>