RadNet, Inc. (RDNT) Earnings

Management highlights

### Core Imaging Business Growth - Broad-based growth driven by strong aggregate and same-center procedural volumes, recent acquisitions, a continued mix shift toward higher-margin advanced imaging, and incremental digital health sales. Strong demand for advanced imaging (MRI, CT, PET/CT) is supported by industry trends and capital investments to expand center capacity. - 25% of current PET/CT volume now comes from high-growth PSMA prostate and brain amyloid studies, up from negligible volume 2-3 years ago. The company is also growing specialty cardiac CCTA programs to drive CT volume growth. - 36% of centers are now operated via health system joint ventures, with a healthy pipeline of new partnership opportunities. Hospitals face growing financial and staffing pressure, increasing demand for RadNet's outpatient and digital solutions. - The company ended Q2 with a strong liquidity position: $726 million in cash, full availability on a $282 million revolving credit facility, and a net debt to adjusted EBITDA ratio of 1.8x. The company completed an opportunistic debt repricing, cutting interest rates by 25 basis points and raising $250 million in new incremental term loan funding for future growth. - Days sales outstanding (DSO) remains at a near company low of 31 days, one of the best metrics in the industry. ### Digital Health & AI Progress - The Gleamer acquisition completed 5 months ago is fully integrated, performing ahead of plan, with legacy Gleamer ARR on track to exceed $30 million by end of 2026. Total cost synergies from the acquisition are expected to reach $4 million annually in 2027. - The company obtained FDA 510(k) clearance for DeepHealth breast ultrasound AI, which automates lesion detection, characterization, and reporting. The solution demonstrated >98% lesion localization accuracy, 8% improved breast cancer detection sensitivity, and 37% lower radiologist interpretation time, and is now commercially available in the U.S. - RadNet now holds 27 FDA clearances and 26 CE marks for AI and enterprise imaging solutions, covering over 100 clinical findings across routine and advanced imaging. - Automated AI-powered draft reporting for radiology has delivered strong early results: over 90% of automated draft reports for thyroid ultrasound are accepted by radiologists without edits, driving 20-30% productivity gains in reporting time and 33% lower sonographer scan time. The company targets 15% of RadNet volumes running through AI auto-draft by end of 2026, rising to over 50% by Q2 2027. - Full enterprise-wide deployment of the DiagnosticSuite AI-native platform is on track for completion by Q1 2027, which will replace the third-party Nuance PowerScribe solution and drive additional cost savings and productivity gains. Automated patient registration is ready for scaling after successful pilots, expected to drive front office productivity gains in the next 6 months.

Guidance

- **Imaging Center Segment**: Management increased 2026 full-year guidance on all key metrics: total net revenue guidance raised to $2.37 billion – $2.42 billion (a $15 million increase at both the low and high end of the prior range); adjusted EBITDA guidance raised to $345 million – $358 million (a $5 million increase at both ends); free cash flow guidance raised to $115 million – $125 million (a $3 million increase at both ends). Capital expenditure guidance remains unchanged at $165 million – $175 million, while cash interest expense guidance was increased by $3 million at both ends to $48 million – $53 million to reflect new incremental borrowings. - **Digital Health Segment**: All 2026 full-year guidance is reaffirmed: revenue guidance maintained at $135 million – $145 million, adjusted EBITDA guidance maintained at $10 million – $12 million. Full-year 2026 ARR is still expected to grow ~91% YoY to over $140 million, with external ARR expected to grow to 65-70% of total ARR by year end. Management reaffirms the long-term target of 20%+ adjusted EBITDA margins for the Digital Health segment by 2027-2028. - **2027 Medicare Reimbursement Initial Analysis**: Initial analysis of the proposed 2027 Medicare Physician Fee Schedule indicates the impact on RadNet's full-year 2027 revenue will be nearly net neutral, with a negative impact of less than $1 million, as proposed RVU increases offset the 1.68% proposed cut to the conversion factor.

Segment performance

1. Imaging Center Segment: Total revenue for the full company (including this segment and Digital Health) was $623 million in Q2 2026, up 25% year-over-year (YoY). The imaging center segment drives the vast majority of total company revenue, with 2026 full-year guidance now set at $2.37 billion to $2.42 billion. In Q2 2026, aggregate advanced imaging procedural volumes increased 21.2% YoY, with same-center advanced imaging volumes up 9.6% YoY: aggregate MRI volumes +21%, same-center MRI +4.0%? Wait correction: transcript notes same-center MRI +4% (typo in original transcript as 40% is inconsistent, likely 4.0%), aggregate CT +20.9%, same-center CT +8.6%, aggregate PET/CT +31.0%, same-center PET/CT +8.8%. Advanced imaging represented 29.9% of total procedural volume in Q2 2026, up 238 basis points from 27.5% YoY. The segment's adjusted EBITDA margin improved 17 basis points YoY to 16.1%. As of Q2 end, 36% (157 out of 442 total centers) are held in health system joint ventures. 2. Digital Health Segment: Q2 2026 revenue was $32.4 million, up 56.5% YoY and 11.4% quarter-over-quarter (QoQ). This splits into $16.1 million AI revenue (up 136% YoY) and $16.3 million enterprise imaging revenue (up 17.3% YoY). Digital Health revenue contributed ~5.2% of total company Q2 2026 revenue. Ending Q2 2026 annual recurring revenue (ARR) was $106 million, up 97% YoY and nearly 9% QoQ, with 63% of ARR coming from external customers outside RadNet. Adjusted EBITDA for the segment was $2.5 million in Q2 2026, up from $1.3 million in Q1 2026 but down from $3.4 million YoY, due to deliberate growth investments.

Risks & headwinds

- Persistent industry-wide labor shortages for technologists and radiologists have created ongoing upward pressure on salaries and wages, which has absorbed some near-term margin improvements from digital efficiency gains and favorable volume mix. Implementation of new digital and AI tools requires ongoing near-term training and integration expenses that suppress near-term margins. - Long backlogs for imaging procedures in high-demand markets can lead to lost patient volume and referrals if capacity is not expanded quickly enough to meet demand. - Forward-looking financial and operational performance is subject to uncertainty, including risks associated with integrating acquired operations, generating expected revenue and EBITDA from acquisitions, successfully recruiting and retaining staff, securing third-party reimbursement for new AI solutions, and growing external digital health sales. Actual results may differ materially from guidance due to these and other risks disclosed in RadNet's SEC filings. - The proposed 2027 Medicare site neutrality provision for hospital outpatient imaging is not yet finalized, and there is no guarantee the final rule will match current proposals, though management expects it will ultimately increase partnership opportunities for RadNet.

Analyst Q&A

• Q: What operational and clinical impact will the newly FDA-cleared breast ultrasound AI have, and what is the timeline for other pending FDA approvals?

  A: RadNet performs 1 million annual breast ultrasound exams, 4x the volume of thyroid ultrasound. The AI automates lesion detection, characterization, and BI-RADS categorization, reduces interpretation and scan time, and is eligible for existing reimbursement via the 0690T CPT code, matching thyroid ultrasound's reimbursement pathway. Full deployment across RadNet's major centers is expected by end of 2026, with full scaling complete by Q2 2027. Pending FDA submissions include mammography image-based risk assessment, MR spine auto-draft, next-generation chest X-ray AI, CT lung AI, and vascular ultrasound, all expected to enter RadNet workflow over the next 2-3 quarters. The company targets 50% of RadNet volumes on AI auto-draft by mid-2027, with 70% of volumes already supported by clinical AI assistance.

• Q: What is driving the strong growth in advanced imaging, and how sustainable is ongoing margin improvement for the core imaging business?

  A: Growth comes from a combination of broader industry trends increasing clinical demand for advanced imaging, internal capacity expansion investments (faster newer MRI scanners, extended scanning hours, and TechLive remote technologists that eliminate room closures from technologist shortages), and new high-growth indications like PSMA PET/CT which now makes up 25% of total PET volume. Digital tools like automated patient registration also improve patient throughput and reduce staffing burdens. While near-term salary inflation and integration costs for recent acquisitions have limited near-term margin expansion, management reaffirms a 100-150 basis point long-term margin improvement target by end of 2028, driven by incremental revenue from reimbursable AI tools and ongoing efficiency gains.

• Q: What is the long-term margin trajectory for the Digital Health segment, and how will margins expand from current levels to the 20%+ long-term target?

  A: The 20%+ adjusted EBITDA margin target for Digital Health by 2027-2028 remains unchanged, and management is more confident in this target after recent strong performance. The near-term lower margin reflects deliberate investment in commercial, service, and implementation headcount to drive scaled growth. Core organic Digital Health business already delivers 30-40% EBITDA margins aligned with standard SaaS business metrics. As acquired businesses like Gleamer are turned profitable and scale increases, overall segment margins will rise toward the long-term target, even as the company continues to invest in new product development.

• Q: What is the expected reimbursement ramp for breast ultrasound AI, compared to the prior thyroid ultrasound launch?

  A: Reimbursement for breast ultrasound AI is expected to ramp faster than thyroid, because payers that already reimburse thyroid AI via the T-code will be easier to convince to cover breast AI for the same indication. Management expects to reach a similar 30-40% reimbursement coverage rate faster than with thyroid. With 4x the volume of thyroid, even at the same coverage rate, incremental revenue from breast ultrasound AI is expected to be 4x that of thyroid at full ramp.