January 30, 2023

Can an AI-supported Radiological Software Tool Objectively Trace Pain Changes?

New Study shows Correlation of Radiographic Alterations to Clinical Outcomes and Synovial Fluid Markers in Knee Joint Distraction (KJD) with IB Lab KOALA

Share this article

Can an AI-supported Radiological Software Tool Objectively Trace Pain Changes?

New Study shows Correlation of Radiographic Alterations to Clinical Outcomes and Synovial Fluid Markers in Knee Joint Distraction (KJD) with IB Lab KOALA

January 30, 2023

“Artificial intelligence in osteoarthritis: repair by knee joint distraction shows association of pain, radiographic and immunological outcome”

Mylène P Jansen, Christoph Salzlechner, Eleanor Barnes, Matthew D DiFranco, Roel J H Custers, Fiona E Watt, Tonia L Vincent, Simon C Mastbergen

New study released in the Oxford Press' "Rheumatology" Journal

The University Medical Center Utrecht recently published a study in the Oxford Press’ “Rheumatology” Journal with the objective to analyze radiographic changes after knee joint distraction (KJD) using IB Lab KOALA, ImageBiopsy Lab’s AI software tool for the assessment of knee osteoarthritis and relate these radiographic alterations to clinical outcome and synovial fluid (SF) markers.

Introduction

Characteristics of knee osteoarthritis (OA) include tissue changes such as cartilage degeneration, osteophyte formation, and subchondral bone sclerosis. IB Lab KOALA is ImageBiopsy Lab’s fully-automated artificial intelligence (AI)-based image processing software device. It is intended to aid medical professionals in the measurement of minimum joint space width, in assessment of the severity of sclerosis, joint space narrowing, and osteophytes based on OARSI criteria and the severity of radiographic knee OA based on Kellgren & Lawrence Grading. Multiple studies have shown that KOALA does not only reduce reading time, but also increases inter-reader agreement.

Knee joint distraction (KJD) is a joint-preserving treatment for relatively young patients with end-stage knee osteoarthritis. Although these patients qualify for total knee transplantation (TKA), their age and the likelihood to outlife the replacement demand for an alternative treatment.

The objective of the current explorative study was to investigate the correlation of AI-detected structural changes with clinical outcomes and synovial fluid markers post operation and to evaluate whether AI- based modules such as IBLab KOALA can assist doctors with the detection and quantification of disease modifying treatments like knee joint distraction.

Rheumatology Key Messages

  1. JSW changes after joint-preserving treatment show significant associations with pain and SF marker changes.
  2. Future trials could consider AI-based measurement methods to generate robust pain-associated imaging analysis results.
  3. These results could help selection for disease-modifying osteoarthritis treatments influencing structural and clinical outcome.

Methods

20 relatively young patients considered for TKA received KJD treatment. The study aimed to find differences in SF markers in the timespan of one year. Inclusion criteria included KL grade ≥2, age <65, no presence of inflammatory joint condition, and no previous or planned joint prosthesis. The study collected SF before, during and after treatment.

KJD treatment was performed using an external device fixed to the femur and tibia using bone pins. The device consisted of two tubes with internal springs that were gradually distracted by 2mm during surgery and 1mm per day to achieve a total of 5mm, confirmed on radiographs. The device was removed after 6-7 weeks.

Posteroanterior (PA) Radiographs were taken before and 1 year after surgery to evaluate the patients' progress.

Patients also filled out digital patient reported outcome measures, including the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) questionnaire, before and after treatment. The primary outcome evaluated in the study was the WOMAC pain subscale (0-100). A summary of the design and follow-up is shown in Fig. 1.

Design and follow-up of the current study. KL: Kellgren–Lawrence; KOALA: Knee Osteoarthritis Labelling Assistant; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index

Radiographic Analysis

ImageBiopsy Lab’s Knee Osteoarthritis Labeling software, KOALA, was used to analyze the radiographs. The results were reviewed by a trained reader - no manual changes were made to prevent bias. The software automatically measures the whole-joint KL grade, medial and lateral joint space narrowing scores, compartmental minimum and standardized JSW, and osteophytes and sclerosis of the medial and lateral tibia and femur. The primary outcome was the most affected compartment's minimum JSW, and the secondary outcome was the most affected compartment's standardized JSW. The ratio between standardized medial and lateral JSW was calculated to define the compartmental imbalance. The most changes were seen in the most affected compartment.

Synovial Fluid (SF) Aspirations

The study collected synovial fluid (SF) samples from the treated knee of the patients at 3 different time points: at the start of the treatment, halfway through the treatment and after the treatment. The samples were analyzed for various biomarkers The assays were done by Mesoscale Discovery or by immunoassay. Two patients were replaced because they had no successful SF aspiration at the baseline.

Statistical Analysis

The Wilcoxon signed rank test and paired t-test were used to analyze changes in categorical and continuous radiographic variables. Linear regression was used to calculate the association between changes in pain and markers, correlation plots were also used. A P-value <0.05 was considered statistically significant and no correction for multiple testing was done.

Results

​​The study evaluated 16 patients and found that roughly half of them had improvements in JSW, KL, and JSN, with the improvement in MAC JSW being the only statistically significant improvement (P < 0.05). The change in MAC JSW was positively associated with changes in WOMAC pain (P < 0.04). Additionally, greater increases in the biomarker MCP-1 and lower increases in TGFβ-1 were found to be significantly associated with changes in MAC JSW (P < 0.05). Furthermore, changes in MCP-1 were also found to be positively associated with changes in WOMAC pain (P < 0.05).

Fig. 2: Summary of results from the current study. The figure shows changes induced by knee joint distraction treatment and possible relations between radiological evidence, SF markers and clinical outcome. Note: the arrows indicate direction of change after knee joint distraction, not the direction of the associations between the different parameters (which is opposite for TGFβ-1: TGFβ-1 in SF increases after knee joint distraction as the arrow indicates, but a lower increase in TGFβ-1 is associated with a greater improvement in joint structure and decrease in pain). JSN: joint space narrowing; JSW: joint space width; KL: Kellgren–Lawrence; LAC: least affected compartment; MAC: most affected compartment; MCP-1: monocyte chemoattractant protein 1; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index

Conclusion

Automatic measurements of joint structure using radiographic techniques revealed that most patients under regular care for KJD (Knee Joint Disease) had improved joint structure. Specifically, the MAC JSW (Mean Articular Cartilage Width at the Joint Space) increased significantly and was found to be related to changes in biomarker levels and even improvements in pain reported by patients.

The implementation of the AI-driven module KOALA has proven to be a valuable asset in optimizing resources, both in terms of time and finances. KOALA enables standardization, mitigating the effects of intra- and inter-reader variability.  Furthermore the results obtained through the use of KOALA have consistently demonstrated a high level of sensitivity to changes, making it a reliable tool for detecting even the slightest variations.

References

Mylène P Jansen, Christoph Salzlechner, Eleanor Barnes, Matthew D DiFranco, Roel J H Custers, Fiona E Watt, Tonia L Vincent, Simon C Mastbergen, Artificial intelligence in osteoarthritis: repair by knee joint distraction shows association of pain, radiographic and immunological outcomes, Rheumatology, 2022;, keac723, https://doi.org/10.1093/rheumatology/keac723