Residual Bone Detection in Boneless Chicken with Dual-Energy X-ray and AI

Boneless chicken is used across retail, foodservice, and further processing. Common formats include chicken breast, thigh meat, fillets, diced chicken, strips, nuggets, and other processed poultry products.

Even after deboning, small residual bone fragments can remain in the product stream. For boneless products, this can become a specification issue, a customer complaint risk, and in some cases a physical hazard concern.

The main challenge is that small bone fragments can be difficult to distinguish from surrounding chicken tissue, especially when they are low-density, irregularly shaped, partially embedded, or presented at different orientations.

RaymanTech provides dual-energy X-ray inspection solutions for residual bone detection in boneless chicken applications. Compared with conventional single-energy X-ray, dual-energy systems provide additional material-related information that can support more challenging inspection tasks. RaymanTech’s dual-energy platform can also be combined with AI-based image analysis and classification for complex inspection applications.

At a Glance

Key Takeaway: Dual-energy X-ray inspection provides additional material-related information for residual bone detection, while AI-based classification supports the analysis of complex and irregular residual bone patterns.

What Is Residual Bone in Boneless Chicken?

Residual bone refers to bone fragments that remain in chicken meat after the deboning process.

Mechanical deboning removes most of the bone, but complete separation is not always possible in high-volume production.

Residual bone may be:

• Small or fragmented

• Irregularly shaped

• Partially exposed

• Embedded in chicken tissue

• Positioned at different orientations

• Different in size and thickness

These variations make residual bone inspection more than a routine visual check.

Why Does Residual Bone Matter?

For boneless chicken, residual bone can affect product quality and specification compliance.

Products sold as boneless are expected to meet the processor’s defined bone-removal requirements. If residual bone remains, the result can be off-spec product, customer complaints, rework, or waste.

From a food-safety perspective, the significance of hard or sharp natural components depends on the product and how it is represented. FDA CPG Sec. 555.425 explains that naturally occurring hard or sharp components may generally be expected in certain foods, but can become an unexpected concern when the product is represented as having that component removed.

For products represented as boneless, processors should therefore evaluate residual bone control together with applicable regulations, product specifications, and customer requirements.

For processors, the practical goal is to detect and remove residual bone according to defined product and inspection requirements.

Why Is Chicken Bone Detection Difficult?

Residual bone is generally more challenging to inspect than many dense foreign materials such as metal or glass.

Bone and chicken tissue can have relatively similar X-ray absorption characteristics, especially when the bone fragment is small or low-density.

Detection can also be affected by:

• Bone size and density

• Bone shape and structure

• Bone orientation

• Product thickness

• Bone position within the meat

• Product overlap

• Natural variation in chicken tissue

A conventional X-ray image mainly shows differences in X-ray attenuation. When the contrast between bone and surrounding product is limited, additional material-related information and advanced image classification can be useful.

Dual-energy X-ray and AI-based classification address these two challenges from different angles.

How Dual-Energy X-ray Supports Residual Bone Detection

Conventional single-energy X-ray uses one energy spectrum to create an image based mainly on X-ray attenuation.

Dual-energy X-ray uses measurements at different X-ray energy levels to provide additional material-related information. This can help distinguish materials with different composition and density characteristics.

For boneless chicken, this matters when the target is a low-density residual bone surrounded by or embedded in chicken tissue.

Dual-energy X-ray can provide additional support when:

• Bone fragments are low-density

• Contrast against chicken tissue is limited

• Products are thick or uneven

• Inspection conditions vary by cut or format

RaymanTech describes its dual-energy X-ray system as using high- and low-energy images for residual bone detection in meats including poultry, pork, beef, and fish, as well as for low-density contaminant inspection and uneven or overlapping products.

Dual-energy is not a guarantee of detection in every case. Performance still depends on the product, thickness, bone characteristics, product presentation, and inspection settings.

That is why application testing remains essential.

How AI Supports Chicken Bone Inspection

Dual-energy X-ray and AI address different parts of the inspection challenge.

Dual-energy provides additional material-related information. AI supports the analysis and classification of complex image patterns.

Residual bone can vary in:

• Shape

• Size

• Thickness

• Orientation

• Position

• Degree of embedding

These variations can make fixed-rule inspection more challenging.

AI-based classification can help analyze and classify potential residual bone patterns while accounting for normal product variation. When combined with dual-energy imaging, AI can use image characteristics together with material-related information to support residual bone inspection.

This is particularly relevant when bone fragments are curved, fragmented, partially embedded, or presented at different orientations.

Dual-energy X-ray provides additional material information. AI adds classification support for complex and irregular targets.

RaymanTech RXD-MS Dual-Energy X-ray Residual Bone Inspection System

RaymanTech’s RXD-MS dual-energy X-ray residual bone inspection system is designed for applications involving residual bone and low-density contaminants. The official product page lists RXD-MS-2416, RXD-MS-4016, and RXD-MS-6030 models and identifies bone detection, thin contaminant detection, low-density contaminant detection, uneven product detection, data transfer, and remote support among the system features.

Key Product Information

RaymanTech’s product selection guide specifically positions the RXD MS C Series for residual bone inspection in bone-in and deboned meat, poultry, and seafood products, while the RXD-MS dual-energy series supports residual bone and low-density contaminant applications.

AI-Based Classification of Irregular Bone Fragments

Residual bone does not always follow a consistent geometric pattern. A fragment may be curved, fragmented, partially obscured by chicken tissue, or positioned differently from other bone fragments.

AI-based classification can help analyze and classify potential residual bone patterns while accounting for variations in shape, size, orientation, and position.

Combined with dual-energy X-ray imaging, this provides an additional classification layer beyond material-related X-ray information for challenging residual bone inspection applications.

Single-Energy vs. Dual-Energy X-ray

For processors evaluating chicken bone inspection technology, the choice between single-energy and dual-energy depends on the actual product and inspection challenge.

Dual-energy should not be viewed simply as a higher-resolution version of conventional X-ray. Its primary value is the additional material-related information generated from different X-ray energy levels.

Where Should Chicken Bone Detection Be Installed?

For many boneless chicken lines, X-ray inspection is placed after deboning and before final packaging or downstream processing.

This allows the processor to inspect the product after the main bone-removal process and before it moves into the next production stage.

The best inspection point depends on the actual process flow. Factors such as product format, packaging, washing, further processing, and line layout should be considered during system selection.

For packaged chicken products, X-ray inspection can also be configured to inspect the finished package, depending on the packaging material and application requirements.

X-ray vs. Metal Detection for Chicken Bone Inspection

For residual bone applications, X-ray inspection is generally more relevant than metal detection alone because X-ray systems can detect certain non-metallic dense materials as well as metal.

The choice between single-energy and dual-energy X-ray should be based on the actual product, target bone characteristics, packaging, line conditions, and inspection requirements.

What Should Processors Consider When Selecting a Chicken Bone Detection System?

The most suitable inspection system depends on the actual application rather than a single equipment specification.

Key factors include:

• Product format

• Product thickness

• Bone size, shape, and density

• Product presentation and overlap

• Line speed

• Packaging

• Inspection technology and algorithm

• Reject system

• Hygienic requirements

• Application test results

For processors handling multiple chicken formats, such as breast, thigh, diced chicken, or formed products, each major product condition should be considered during application evaluation.

Application Testing Matters

There is no single inspection setting that fits every boneless chicken product.

Chicken breast, thigh meat, diced chicken, and formed products can differ in thickness, density, moisture, shape, bone characteristics, and presentation.

RaymanTech recommends application testing with representative samples before equipment selection. The company’s X-ray product pages also direct customers toward sample testing to validate inspection performance before purchase.

Testing can evaluate:

• Residual bone detection

• Irregular bone fragment detection

• Product presentation

• Reject performance

• Target throughput

• AI classification under representative conditions

Representative samples should include normal product, known residual bone samples, typical product variation, and relevant foreign materials.

This provides a practical basis for matching the inspection configuration to the processor’s actual product and production requirements.

FAQ

Industry References